Labour supply and employment in the euro area countries - developments and challenges
-
Upload
independent -
Category
Documents
-
view
1 -
download
0
Transcript of Labour supply and employment in the euro area countries - developments and challenges
Task force of the Monetary Policy Committee of the European System of Central Banks
Occas iOnal PaPer ser i e snO 87 / J une 2008
nO
87
Ju
ne
2008
Occ
asiO
nal
PaP
er s
erie
s
laBOur suPPlY anD
eMPlOYMenT in THe
eurO area cOunTries
DeVelOPMenTs
anD cHallenGes
OCCAS IONAL PAPER SER IESNO 87 / JUNE 2008
Task Force of the Monetary Policy Committee
of the European System of Central Banks
LABOUR SUPPLY AND EMPLOYMENT
IN THE EURO AREA COUNTRIES
DEVELOPMENTS AND CHALLENGES
This paper can be downloaded without charge from
http://www.ecb.europa.eu or from the Social Science Research Network
electronic library at http://ssrn.com/abstract_id=1084913.
In 2008 all ECB publications
feature a motif taken from the €10 banknote.
© European Central Bank, 2008
Address Kaiserstrasse 29
60311 Frankfurt am Main
Germany
Postal address Postfach 16 03 19
60066 Frankfurt am Main
Germany
Telephone +49 69 1344 0
Website http://www.ecb.europa.eu
Fax +49 69 1344 6000
All rights reserved. Any reproduction, publication or reprint in the form of a different publication, whether printed or produced electronically, in whole or in part, is permitted only with the explicit written authorisation of the ECB or the author(s).
The views expressed in this paper do not necessarily refl ect those of the European Central Bank.
ISSN 1607-1484 (print)
ISSN 1725-6534 (online)
3ECB
Occasional Paper No 87
June 2008
CONTENTS
EXECUTIVE SUMMARY AND CONCLUSIONS 6
1 INTRODUCTION 10
2 CONCEPTUAL ISSUES RELATING TO
LABOUR SUPPLY AND THE MACROECONOMY 12
3 MAIN TRENDS IN LABOUR SUPPLY 15
3.1 Population and the labour force 16
3.2 Participation rates 18
3.2.1 Participation rates by
individual characteristics in
the euro area 18
3.2.2 Participation rates in the
euro area countries 20
3.2.3 Compositional effects in
participation 21
3.2.4 Cohort effects in participation 22
3.3 Hours worked 27
3.3.1 Hours worked in the euro
area and in the United States 27
3.3.2 Hours worked in the euro
area by different worker groups 29
3.3.3 Hours worked per week in
the euro area countries 30
3.3.4 Compositional effects in
hours worked 31
3.4 Immigration 32
3.5 The supply of knowledge and skills 38
3.5.1 Educational attainment 38
3.5.2 The quality of educational
attainment 40
4 STRUCTURAL POLICIES AND THEIR
EFFECT ON LABOUR SUPPLY 45
4.1 Tax and benefi t systems 46
4.1.1 Unemployment benefi t
systems 46
4.1.2 Labour taxes 47
4.1.3 Pension systems 52
4.2 Parenthood and participation 56
4.2.1 Fertility and female
participation: what is the
relationship? 56
4.2.2 Availability and cost of
child care 57
4.2.3 Maternal and parental leave
opportunities 59
4.2.4 Part-time work
opportunities 59
4.3 Immigration Policies 61
4.3.1 Policies affecting
immigrants once in a country 63
4.4 Education systems and incentives 64
5 MATCHING LABOUR SUPPLY AND DEMAND 71
5.1 Returns to education 71
5.2 Relative unemployment rates of
different groups of workers 74
5.2.1 Institutions affecting the
matching of labour supply
with labour demand 74
5.2.2 Educational mismatch 76
5.2.3 Regional mismatch 79
5.2.4 Mismatch by age, gender
and nationality 79
5.3 The future demand for labour 83
ANNEX 1
MEASUREMENT ISSUES 85
ANNEX 2
DATA SOURCES AND DESCRIPTIONS 87
ANNEX 3
DETAILED TABLES AND CHARTS 91
ANNEX 4 108
REFERENCES 110
EUROPEAN CENTRAL BANK OCCASIONAL
PAPER SERIES SINCE 2007 121
LIST OF BOXES:
Box 1 Projections of labour supply in the
euro area and euro area countries 25
Box 2 Recent experiences with
immigration in euro area countries 35
Box 3 Human Capital and growth 40
Box 4 Labour quality growth in the
euro area and euro area countries 42
Box 5 Recent reforms of tax and benefi t
systems. Case study: Ireland and
Germany 50
CONTENTS
4ECB
Occasional Paper No 87
June 2008
Box 6 Enhancing the participation rate of
older workers: the need for a
comprehensive strategy 53
Box 7 Reform of childcare and fl exible
working arrangements. Case study:
France and Finland. 60
Box 8 Incentives for fi rms and workers to
invest in human capital
accumulation and life-long
learning 68
Box 9 Returns to education in the
euro area countries 72
Box 10 The gender pay gap and
discrimination 82
Box 11 ILO defi nition of unemployment 85
Box 12 Cross-border commuting in the
euro area. Case study:
Luxembourg 108
JEL classifi cation: E5, J1, J2, J6
Keywords: Labour supply, employment, participation, hours worked, immigration, skill and
education, structural policies, labour demand, unemployment, euro area countries, labour markets,
taxes and benefi ts, childcare, pensions, training, human capital, labour quality, working time and
contracts, discrimination, mismatch, returns to education.
5ECB
Occasional Paper No 87
June 2008
TASK FORCE OF THE
MONETARY POLICY
COMMITTEE OF THE
EUROPEAN SYSTEM
OF CENTRAL BANKS
TASK FORCE OF THE MONETARY POLICY COMMITTEE OF THE EUROPEAN SYSTEM OF CENTRAL BANKS
This report was drafted by an ad hoc Task Force of the Monetary Policy Committee of the European
System of Central Banks. The Task Force was chaired by Klaus Masuch. The coordination and
editing of the report was carried out by Jarkko Turunen (during the period March-August 2007)
and Melanie Ward-Warmedinger (during the period August 2007-June 2008). Clare Childs, Dung
Hoangkim and Stefanie Peuckmann provided editorial assistance.
The full list of members of the Task Force is as follows:
Klaus Masuch European Central Bank
Ramon Gómez-Salvador European Central Bank
Nadine Leiner-Killinger European Central Bank
Rolf Strauch European Central Bank
Jarkko Turunen European Central Bank
Melanie Ward-Warmedinger European Central Bank
Jan De Mulder Nationale Bank van België/Banque Nationale de Belgique
Harald Stahl Deutsche Bundesbank
Yvonne McCarthy Central Bank and Financial Services Authority of Ireland
Daphne Nicolitsas Bank of Greece
Aitor Lacuesta Banco de España
Mathilde Ravanel Banque de France
Piero Cipollone Banca d'Italia
Christelle Olsommer Banque centrale du Luxembourg
Alfred Stiglbauer Oesterreichische Nationalbank
Álvaro Novo Banco de Portugal
Klara Stovicek Banka Slovenije
Heidi Schauman Suomen Pankki
Other contributors:
Almut Balleer European Central Bank
Kieran McQuinn Central Bank and Financial Services Authority of Ireland
Pasqualino Montanaro Banca d'Italia
Alfonso Rosolia Banca d'Italia
Eliana Viviano Banca d'Italia
Cláudia Filipa Duarte Banco de Portugal
Matija Vodopivec Banka Slovenije
6ECB
Occasional Paper No 87
June 2008
EXECUTIVE SUMMARY AND CONCLUSIONS
RATIONALE AND MAIN OBJECTIVE OF THE REPORT
The functioning of labour and product markets
affects the economic environment in which
monetary policy is conducted. For example,
structural policy measures which enhance
labour supply and employment growth increase
the pace at which an economy can grow
without higher infl ation. A greater fl exibility
of euro area labour markets and wages would
reduce adjustment costs and infl ation pressures
in the case of adverse supply shocks and
augment resilience of the economy, facilitating
the conduct of the stability-oriented monetary
policy of the ECB. In addition, an enhanced
fl exibility of wages and labour mobility is
needed to limit employment losses in the case
of adverse country specifi c shocks, thereby
facilitating the functioning of EMU.
Labour markets in the euro area, as in other
developed economies, face the triple challenge of
demographic change, technological progress and
globalisation. Demographic change (a decrease
in the size of the working age population and an
increase in the average age of the labour force)
reinforces the need to increase labour market
participation and employment in the euro
area. Increasing the share of people working
will help to support the euro area’s potential
output and per capita income, and reduce the
old-age dependency ratio. All this would help
to fi nance pension and health care systems
and to reduce the per capita fi nancing burden
for those who have to pay for these systems
via taxes and social security contributions.
Technological progress results in fi rms
searching for people with new types of skill and
knowledge. In order to support both high wages
and low unemployment, it is important that the
population of the euro area is well educated
and trained in the types of skills that fi rms
seek and that workers invest in enhancing and
developing their skills over the course of their
working lives. Effi cient schooling and education
systems, including vocational training, will play
an important role in enhancing investment in
education and equipping individuals (including
older persons) with both the skills demanded in
the labour market at any point in time and the
general competences that will allow them to
adapt fl exibly to new developments in labour
demand. In this context, training in skills sought
by traditional industries also remains important.
Finally, globalisation increases the ease with
which fi rms can either hire labour from abroad
(through immigration) and/or relocate their
production and services. It also increases the
competition faced by fi rms in the production
of goods and services. For workers in Europe,
this means that they have to remain competitive
in terms of the interplay between type and
level of skills, adaptability, productivity and
compensation packages.
Against this background and to successfully face
this triple challenge, conditions must be in place
that enhance the quantity and quality of labour
supply and effi ciently match the workforce with
fi rms’ demand for labour. This is necessary to
maximise individuals’ income and welfare and –
at the aggregate level – an economy’s potential
output, allowing an increase in the economy’s
rate of sustainable growth. Well-designed and
fl exible labour and product market institutions
are essential to this process. Moreover, structural
policy changes that enhance incentives for
schools, universities and fi rms to identify and
develop the “right” skills are needed. In addition,
the euro area should make the best use of global
labour supply through immigration, ensuring
that immigrants are effectively integrated into
its labour market and society.
The aim of this report, which has been prepared
by a Task Force of the Monetary Policy
Committee of the Eurosystem, is to describe
and analyse the main developments in labour
supply and its determinants in the euro area,
review the links between labour supply and
labour market institutions, assess how well
labour supply refl ects the demand for labour in
the euro area and identify the future challenges
for policy-makers. The data available for this
report generally cover the period from 1983 to
spring 2007. The cut-off date for the euro area
statistics included was 14 December 2007.
7ECB
Occasional Paper No 87
June 2008
EXECUTIVE
SUMMARY AND
CONCLUS IONSMAIN FINDINGS ON DEVELOPMENTS IN
LABOUR SUPPLY:
Following a conceptual introduction to labour
supply from a macroeconomic perspective in
Chapters 1 and 2, Chapter 3 documents the key
developments in labour market participation,
employment and hours worked in the euro area
and EU Member States. Particular emphasis is
given to issues relating to the quality of labour
and human capital and immigration. Chapter 3
fi nds that:
i) over the period 1996-2007, overall labour
market developments appear to have been
quite favourable for the euro area as a whole.
Employment growth accelerated signifi cantly,
with the equivalent of 21.6 million new jobs
created. This contributed to a substantial
3.9 percentage point reduction in the
unemployment rate, which stood at 7.5% in
spring 2007. As a result, the positive gaps in
labour market participation and employment
rates between the United States and the euro area
became narrower and are now closer to those
prevailing in the early 1970s. In the euro area,
the total labour market participation rate rose by
5.6 percentage points over the period 1996-2007,
to nearly 71%, and the employment rate rose by
7.7 percentage points to over 65%. The highest
levels of participation in 2007 were registered
for men at 78% (employment at 73%), for prime-
aged individuals at 85% (79%), for citizens of the
new EU Member States (countries joining the
EU since 2004) at 77% (69%) and for the highly
educated at 88% (85%). Increases in participation
have compensated for a slight decline in the
working age population in recent years. However,
whilst much has been achieved, there is no room
for complacency. For example, labour supply
projections show that even if these positive
participation trends continue, they will soon no
longer be suffi cient to counteract the fall in the size
of the working age population. Furthermore, by
international standards, many euro area countries
continue to exhibit high unemployment rates and
low labour market participation;
ii) labour supply composition has changed over
time. Particularly women and the so-called
non-EU15 immigrants (immigrants from the
new EU Member States and non-EU countries)
have entered the labour market in increasing
numbers, and older workers have tended to
remain in the labour market for longer. The
participation of these groups increased by
9.0 percentage points, 7.4 percentage points
and 10.5 percentage points respectively
over the period 1996-2007. Changes in
educational levels, preferences and social
norms have, over time, played a role in
increasing female labour market participation
(through so-called cohort effects). Experiences
with immigration vary considerably by
country, but on the whole, immigrants have
contributed positively to labour supply and
employment, enhancing competition in labour
markets and helping to fi ll skills shortages.
Cross-border commuting within the euro
area has increased threefold in the last ten
years. Potential for further increases in labour
market participation exists, particularly among
younger workers (following completion of
their education), women and older workers, and
through enhancing immigrants’ integration;
iii) these developments have been accompanied
by an increase in labour quality and the
share of workers with higher education,
particularly those with tertiary level education
(by 6 percentage points between 1996 and 2007,
to a level of 20.7%). This implies an increase
of 16.7 million in the number of people with
tertiary level qualifi cations. The proportion of
low-skilled workers decreased in the euro area
over the period considered (by 9 percentage
points to a level of 37.7%). However, the level
of human capital in some euro area countries
lagged behind that in other advanced countries,
and labour quality growth seems to have slowed
towards the end of the 1990s, highlighting
the need for further increases in educational
attainment and on-the-job training; and
iv) as the number of workers has increased,
the average hours worked per week has
declined in the euro area (by a total of
1.2 hours over the period 1996-2007). This
refl ects changes in working time regulations
8ECB
Occasional Paper No 87
June 2008
and especially an increase in part-time jobs,
which has helped women, in particular, to join
the labour market in greater numbers. The ratio
of part-time to total employment increased
by around 5 percentage points over the
same period.
MAIN FINDINGS ON THE STRUCTURAL POLICIES
DETERMINING PARTICIPATION AND EMPLOYMENT
AND POLICY CONCLUSIONS:
Chapter 4 reviews how structural policies affect
labour supply and may have contributed to the
developments observed. Chapter 5 presents
evidence on how well the characteristics of
labour supply have refl ected those of the
demand for labour over the last two decades and
discusses how to respond to the main challenges.
The welcome developments in participation and
employment rates over the last decade partly
refl ect increased labour market fl exibility and
reform progress. However, this progress has been
quite uneven across countries. Looking forward,
a priority for economic policies is to ensure high
potential output, greater fl exibility and resilience
of the euro area countries to shocks. While there
is often no single solution for all, it is important
that countries learn from each other and from
best practices to develop new and better labour
market institutions and policies which will help
to achieve desired results. Identifying which
economies have performed best with regard to
which specifi c policy areas should help with
this. Furthermore, it is important to consider
reform packages as a whole when embarking
on reform processes, in order to predict trade-
offs in a timely manner and ensure that policies
are complementary within the framework of a
comprehensive reform strategy. The key fi ndings
and policy conclusions of the report are as
follows:
i) The analysis undertaken in Chapter 4 shows that
structural policies affect, inter alia, individuals’
labour market decisions and household income.
There is a need to further optimise policies in the euro area and to increase the labour market participation and employment of all groups, especially females and non-prime-aged
workers. Reducing high marginal tax rates and
tax wedges, high unemployment benefi t levels
and long durations, weak work availability
requirements and early retirement schemes
would stimulate labour supply and income.
Tax and benefi t systems should not discourage
older workers from voluntarily staying longer
in the labour market (for example, fi nancial
disincentives for extending participation beyond
standard retirement ages should be removed).
Reducing taxes and social security contributions
on labour, especially if funded by expenditure
constraint and enhanced effi ciency of public
fi nances, would enhance employment and net
wages, as well as provide additional incentives
for individuals to move from unemployment to
work or to invest in human capital. Restrictions
on working arrangements and labour contracts
should be reduced to allow for more fl exible
working hours (both higher and lower
than standard contracts) and more fl exible
employment protection. Furthermore, so called
“work/family reconciliation” policies that
facilitate the combination of work and a family
(such as the provision of affordable childcare,
parental leave and part-time work opportunities)
should be well-designed to support and
encourage labour market participation and
ensure equal opportunities. Evidence suggests
that the success of reform measures hinges on
the use of comprehensive reform strategies that
take into account the factors that may infl uence
individuals’ decisions to work and the ease with
which they fi nd a job.
ii) Chapter 5 shows that labour market regulations and institutions need to be more fl exible to facilitate the matching of labour supply with labour demand (across skills,
worker groups, sectors and regions). Labour
market institutions should be geared towards
lowering adjustment costs (e.g. by reducing
employment protection) and barriers to cross-
occupation and geographical labour mobility
(e.g. through common educational standards
and easier pension transfers) and should offer
appropriate incentive-compatible fi nancial
support for those temporarily unemployed and
job search assistance. Flexible wage bargaining
9ECB
Occasional Paper No 87
June 2008
EXECUTIVE
SUMMARY AND
CONCLUS IONSis important for allowing wages to refl ect local
labour market conditions (such as regional and
skill-specifi c unemployment rates, regional and
sectoral productivity growth and workers’ skills).
Wages that are not suffi ciently differentiated –
for example, by skill or region – exacerbate
the mismatch between labour supply and
labour demand (also by not providing the
incentives for capital to shift to areas with
high unemployment), thus increasing the
unemployment rates of certain skill groups and
regions. Institutions have an important role
to play in matching labour supply with labour
demand. Public employment services need to
be more effi cient. Institutional arrangements
that hinder the employment of low-skilled
workers (such as excessive minimum wages)
should be avoided. Contractual arrangements
should give individual workers greater freedom
to agree on contract details (e.g. working hours
and options to invest in additional training).
Moreover tight product market regulation,
and more generally measures that hinder or
restrict competition, are an impediment to job
creation. Policies that increase competition in
goods and service markets, such as those that
reduce the administrative burden on fi rms start-
ups and remove statutory barriers to entry in
certain sectors, would help support employment
creation.
iii) Chapters 3, 4 and 5 highlight the need to increase skills and knowledge (and thus the quality of labour supply) and the transferability of skills. Employment normally enhances skills
because workers learn by doing. Therefore,
bringing unemployed or inactive people into
jobs will, over time, enhance individuals’ labour
productivity and thus real wages. An effi cient
framework for training and counselling the
(long-term) unemployed should help them
to remain employable. The large differences
in tertiary education funding between the
United States and the euro area should be
addressed also by enhancing conditions for
private funding. In addition, the labour market
(including fi rms) should play a stronger role in
signalling to education systems and workers
which skills are expected to be in short supply.
Good quality education should be ensured, in
particular, by enhancing the relevant incentives
and recognition for young people, workers and
fi rms to invest in education and training. The
effi ciency and service orientation of education
institutions should be improved, e.g. by
enhancing some elements of competition and
external quality controls.
iv) Evidence presented in Chapters 3, 4 and
5 also shows that the euro area should make better use of skills from outside the euro area.
However, immigration is not a substitute for
economic reform geared towards removing
barriers to high labour market participation.
Rather, it provides a means through which
the euro area can tap into the global supply
of labour resources and fi ll domestic skill
shortages. From an economic perspective,
the benefi ts derived from immigration depend
on the characteristics of migrants entering
the euro area and the ease with which they
integrate into work and society. Immigration
policy should be closely aligned with the
skills needed by the labour market, ensuring
appropriate migration fl ows that have the
potential to fi ll gaps in labour supply and ease
adjustments over time (particularly in view
of population ageing). Selective migration
policies which limit labour mobility within
the EU should be avoided and replaced
by measures that support labour market
mobility (such as the increased portability
of pension rights). Policies that ensure the
successful integration of immigrants into
the active workforce and society as a whole
(e.g. through incentives to broaden language
skills) are crucial. Successful integration
is also important because some migratory
fl ows are not steered in line with immediate
skill needs (such as family reunifi cation and
asylum seekers). In some countries, reforms
may be necessary before larger numbers of
immigrants can be expected to be integrated
successfully into labour markets.
10ECB
Occasional Paper No 87
June 2008
1 INTRODUCTION 1
This report aims to analyse the main
developments in labour supply and its
determinants in the euro area, describe the level
and structure of employment, 2 review some of
the links between labour supply, labour market
institutions and economic performance, assess
how well labour supply refl ects the demand for
labour in the euro area, and identify challenges
relating to these topics for policy-makers.
These issues are very important for economic
development and welfare more generally, both
at an individual and aggregate level. They are
also important from a central bank perspective,
since labour supply affects the environment in
which monetary policy is conducted. The precise
impact of changes in labour supply on potential
and actual output, as well as on the natural and
actual unemployment rate, is affected by the
design and fl exibility of labour and product
market institutions and by the adjustment of
labour costs to these developments. In this
report labour supply is understood in a broad
sense to cover the size and composition of the
labour force (including the self-employed) by
age, gender, educational attainment level and
nationality (as an indicator of immigrant status)
and is thus analysed in terms of both quality and
quantity.
The main developments in euro area labour
markets, from a long-term perspective, are a
signifi cant decline in employment rates and an
increase in unemployment rates, which started
in the 1970s. Over the most recent decade
developments reversed, leading to rising labour
force participation and employment rates, as
well as declining unemployment rates. Policy
initiatives, such as the European Employment
Strategy and the Lisbon Agenda for Growth
and Jobs, and a favourable macroeconomic
environment, have supported these recent
improvements. Many euro area countries have
made some progress with labour market reforms,
such as improving work incentives, and by
reducing product market regulations, which also
affect labour market performance. However, this
progress has been quite uneven across countries
and further reform of labour and product markets
is needed. By international standards, most euro
area countries still have high unemployment
rates and low labour market participation.
Such levels cannot be explained by factors like
cyclical development, suggesting that there
are still structural and institutional barriers to
labour supply and employment within the euro
area. In this respect, it is possible to identify a
number of specifi c labour supply characteristics
in the euro area (and, more widely, Europe)
that warrant further consideration. These
include low (but increasing) female labour
supply rates, a relatively low youth labour force
participation rate but an increasingly educated
workforce, relatively high early retirement rates,
recent increases in immigration, high youth
unemployment rates and the existence of labour
market mismatches across certain groups of the
labour force, regions and skills.
This report is organised as follows: Chapter 2
briefl y discusses labour supply in the
macroeconomy from a conceptual point of view.
It details how developments in labour supply are
relevant to economic developments, welfare and
monetary policy, outlining the possible impact
of changes in both the quantity and quality of
labour supply on wages and output over the
short to medium term. Institutions’ conceptual
role in shaping this impact is also discussed.
Chapter 3 presents empirical evidence on the
main trends in the quantity and quality of labour
supply and employment in the euro area and
euro area countries over the last two decades.
It considers how the age, gender, educational
attainment and nationality profi le of labour
supply in the euro area has changed over time
by assessing the participation and employment
of these sub-groups. Particular emphasis is
given to developments in immigration and the
supply of human capital. Chapter 4 reviews
how structural policies affecting labour supply –
namely tax and benefi t systems, work-family
Prepared by J. Turunen and M. Ward-Warmedinger.1
Providing a comprehensive analysis of the determinants of 2
employment would require a deeper analysis also of labour
demand issues and falls outside the scope of this report.
11ECB
Occasional Paper No 87
June 2008
I INTRODUCTION
balance policies (including the provision of
childcare and part-time work opportunities),
immigration and educational policy – may have
helped shape the labour supply developments
of the sub-groups considered in the previous
chapter. It presents qualitative assessments of
key reforms and structures and their impact
on the quantity and quality of labour supply.
Looking forward, policies affecting labour
supply will need to accommodate a number of
challenges. Chapter 5 presents evidence on how
well labour supply has refl ected the demand for
labour over the last two decades by analysing
the returns to education and the level and
change in unemployment rates by skill, region,
age, gender and nationality. It presents a brief
overview of how labour and product market
institutions affect the matching of labour supply
with labour demand. Finally, it discusses the
implications of globalisation, demographic and
technological change for the future composition
of labour supply.
12ECB
Occasional Paper No 87
June 2008
2 CONCEPTUAL ISSUES RELATING TO LABOUR
SUPPLY AND THE MACROECONOMY 3
Labour supply developments, in terms of size,
quality and composition, are a major determinant
of an economy’s potential output, affecting an
economy’s rate of sustainable growth. Increasing
the share of people and skills in work will also
help support per capita income and reduce the
old age dependency ratio, which would help
reduce the fi scal burden related to population
ageing. Furthermore, the cyclical sensitivity of
labour supply can affect labour market tightness
and thereby infl uence the outlook for wages
and prices and infl ation dynamics over the
business cycle frequency. The labour supply’s
precise impact on potential and actual output,
and on the natural and actual unemployment
rate, is affected by the design and fl exibility of
labour and product market institutions and the
response of labour cost developments. Sub-
optimal structural and fi scal policy measures
may undermine productivity and increase
structural unemployment, with consequences
for monetary policy. Such factors also affect
individuals’ decisions to supply labour and
the types of labour supplied (discussed further
in Chapter 4). Changes in the composition of
labour supply 4 relative to demand can have
important consequences for the unemployment
rates of particular groups of workers (and
thus total unemployment rates), especially if
labour markets or wages are not suffi ciently
fl exible (discussed further in Chapter 5). These
factors infl uence labour supply developments’
impact on actual and expected wage and price
pressures, with possible implications for the
conduct of monetary policy. This chapter briefl y
reviews how developments in labour supply
affect the macroeconomy and their relevance
for monetary policy.
Labour supply is a key contributor to
economic growth. Both an increase in labour
input as measured by total hours worked
(employment times hours worked per worker,
sometimes referred to as labour utilisation)
and improvements in human capital have the
potential to contribute positively to real GDP,
income growth and welfare. Labour utilisation
is largely determined by developments in
population growth (itself a function of fertility
and mortality rates and immigration, discussed
further in Box 1) and the likelihood that those
in the working age population participate in the
labour market (dependent on job opportunities
and incentives/disincentives to enter the labour
market, discussed further in Chapters 4 and 5).
Furthermore, hours worked from a long-run
perspective – that is total hours of work over an
individual’s lifetime – affect the level of output.
The potential for the supply of human capital
(encompassing factors such as the quantity
and quality of formal education, labour market
experience and on-the-job training, as well as
a broader set of competencies, e.g. cognitive
abilities) to contribute to growth appears
substantial as refl ected in the prominent role
of human capital in modern growth theory.
In particular, endogenous growth models
suggest that improvements in human capital
can generate technological progress and
thus productivity growth in the long term.5
Human capital contributes to measured total
factor productivity growth through changes
in the skill composition of the workforce and
possible interactions between human capital and
technology adoption.6
Changes in labour supply at the business cycle
frequency can affect labour market tightness.
Besides the magnitude and the persistence of a
positive or negative labour supply shock, the
extent and propagation of its impact on
macroeconomic variables depends on its
interaction with the economy’s institutional
framework. Three main categories of rigidities
infl uence the transmission of a labour supply
shock (and on the short-run unemployment-
infl ation trade-off), namely real wage rigidity,
the rigidity of contracts (e.g. contract duration
Prepared by K. Stovicek, J. Turunen and M. Ward-3
Warmedinger.
See Section 3.2.3 and section 3.2.4 for a discussion of 4
composition effects in labour market participation and their
effect on aggregate levels and trends in labour supply.
Barro and Sala-i-Martin (2004).5
See Gomez-Salvador et al. (2006) for a more detailed description 6
and further references.
13ECB
Occasional Paper No 87
June 2008
2 CONCEPTUAL ISSUES
RELATING TO LABOUR
SUPPLY AND THE
MACROECONOMY
and contract design fl exibility) and rigidities in
the legislative and economic framework
(e.g. employment protection).7 The interaction
of a positive labour supply shock with rigid
labour market institutions may shift adjustment
over the business cycle from prices (wages) to
quantities, thus increasing unemployment in
particular in the short to medium run.
Furthermore, lower incentives to take up a job
(e.g. stemming from disincentives created by
the unemployment insurance system), as well as
regulations restricting labour demand, such as
high minimum wages, may contribute further to
a larger accommodation of increased labour
supply via unemployment. From this perspective,
labour and product market reforms (that, for
example, reduce red tape, enhance real wage
fl exibility, lower employment protection,
increase incentives from unemployment
insurance systems etc. – see Chapter 4) may
contribute to an economy’s shock absorption
capacity, by either dampening the unemployment
effect of the labour supply shock, or speeding
the economy’s return to a high employment
equilibrium by lowering the persistence of
employment and output fl uctuations.8 Some of
these structural reforms may also help lower the
natural rate of unemployment.
The measured participation rate tends to be pro-
cyclical, since persons on the edge of labour
market attachment (who have acquired little
work experience or career-specifi c education 9)
and who are more likely to move in and out of
the labour market react to changes in labour
market conditions and job availability (discussed
further in Chapter 4). The degree to which labour
supply adjusts through participation is also
affected by market rigidities, job availability
and matching frictions. Movements in and out
of the labour force may complicate the task
of measuring labour market participation and
unemployment with precision, since changes in
the measured participation rate may not refl ect
a change in individuals’ preferences (since, for
example, labour market rigidities may mean
that those wishing to work are discouraged
from participating in the labour market).
Issues related to the diffi culty of measuring
labour market participation and unemployment
(arising from, inter alia, actual moves in and
out of the labour market or from work in the
informal economy) are discussed further in
Annex 1 and Box 11. Average hours worked
(per person employed) are also procyclical
since they provide an adjustment mechanism
for employment. This form of adjustment may
be particularly relevant for fi rms expecting a
change in product demand to be short-lived and
operating in the labour market with substantial
adjustment costs to employment (in terms of
persons employed). The shift from full-time to
part-time employment, and vice versa, is another
driver of changes in hours worked.
Changes in the composition of labour supply
relative to demand also have important
consequences for the wage structure and
unemployment rates of sub-groups of workers.
Assuming unchanged demand for each labour
type and fl exible wages, an increase (decrease)
in the relative supply of a specifi c type of
workers results in a lower (higher) wage for
that type, relative to other types. If, however,
relative wages are not completely fl exible, the
change in relative supply results in changes in
relative unemployment.
Finally, changes in labour supply naturally have
implications for monetary policy, since they
may affect the actual and natural rates of
unemployment and infl ation dynamics. The
institutional framework is also important in this
context, since it can affect both the transmission
of a labour supply shock to infl ation dynamics
(in particular through its affect on wage
developments and their pass-through to prices)
and the transmission of monetary policy to
infl ation itself (through its effect on adjustment
in the labour market). Recently, the development
See Layard et al (2005).7
As stated in Duval, Elmeskov and Vogel (2007), there is no 8
simple link between rigidities in labour and product markets and
resilience, since institutions that dampen the initial impact of a
shock may also increase its persistence and vice versa. Therefore
the net effect of structural policies remains an empirical issue.
See e.g. Aaronson, Fallick, Figura et al. (2006), Bradbury 9
(2005), Elmeskov and Pichelman (1993), Clark et al (1979).
14ECB
Occasional Paper No 87
June 2008
of micro-founded structural models to
incorporate various labour market features
shows that labour market rigidities, notably
wage rigidity, increase infl ation persistence,
thereby changing the short-run trade-off between
infl ation and unemployment.10 In this framework,
labour market rigidities may affect an optimal
monetary policy aiming to reduce the welfare
costs of macroeconomic fl uctuations.11 In a more
fl exible labour market, wages could be expected
to more closely refl ect workers’ marginal
productivity, reducing the impact of a labour
supply shock on short-term infl ation.
Within the literature to date, labour supply shock transmission is 10
strongly dependent on a model’s characteristics.
Blanchard and Gali (2007) explore the transmission of a 11
productivity shock in the framework of real wage rigidity. They
fi nd that to the extent this transmission has implications in the
medium term, optimal monetary policy may take into account
both infl ation and the output variability. Christoffel, Kuester
and Linzert (2006) fi nd that in the presence of labour market
frictions and wage rigidity, an optimal monetary policy might
consider both wage and price infl ation in their monetary policy
reaction function.
15ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY3 MAIN TRENDS IN LABOUR SUPPLY12
This chapter presents empirical evidence on the
main trends in the quantity and quality of labour
supply in the euro area from the early 1980s
onwards using data from Eurostat’s Labour Force
Survey (EU-LFS).13 Labour supply developments
are discussed in terms of changes in population,
labour market participation and hours worked by
four main individual characteristics: age, gender,
education and nationality. The chapter also briefl y
covers measurement issues relating to the main
indicators of labour supply, with emphasis on the
measurement of human capital and the quality
of labour supply. Finally, this chapter describes
recent developments in immigration with a focus
on the euro area countries where migration has
been particularly important for labour supply.
When using data on employment, unemployment
and the labour force, it is important to take note
of some measurement issues that can potentially
lead to some mis-measurement of the true levels
of these variables. For instance, a number of
individuals working in the shadow economy and
in household production are not registered as
employed. Moreover there are individuals who do
not work but are available to work, and may not
be captured by the defi nition of unemployment
used to collect labour market statistics in the
EU-LFS, resulting in an underestimation of the
actual labour supply (see Annex 1 for a more
detailed discussion). For example, since non-
participation and the number of hours worked
are not just the result of individual preferences,
but also of market rigidities which undermine
labour demand, “true” unemployment in the
euro area may be higher than measured.14 Taking
these issues into account in a systematic way is
beyond the scope of this study. However, given
these potential measurement problems regarding
the distinction between non-market activities,
inactivity and unemployment, and the overall
importance of employment, this report also
presents information on employment rates.
Main fi ndings of this chapter include an
increase in labour market participation of
5.6 percentage points in the euro area over the
period 1996 to 2007, with the highest levels of
participation in 2007 registered for men (78%),
prime-aged individuals (85%), non-nationals
from the 12 new Member States (77%) and the
highly educated (88%). In particular, women
and non-EU15 immigrants have entered the
labour market in increasing numbers, and older
workers have tended to stay in the labour market
longer. The labour market participation of these
groups increased by 9.0 percentage points,
7.4 percentage points and 10.5 percentage
points respectively (0.9 percentage point,
0.6 percentage point and 0.9 percentage point
on average per year). Cohort effects linked
to changes in educational levels, preferences
and social norms over time played a role in
increasing female labour market participation.
Over this period, developments in total hours
worked in the economy, as an alternative
measure of labour input, show an upward trend
similar to that observed for employment. At the
same time, average weekly hours of work per
employed person in the euro area have declined
by 1.2 hours per week over the period 1996 to
2007, largely due to the increase in part-time
jobs, which increased by around 5 percentage
points as a share of total employment. These
developments have been accompanied by an
increase in the share of the population with
higher education, in particular those with
tertiary level education (by 6 percentage
points since 1996, to 20.7% in 2007). The
proportion of the population with a low level
of education has fallen (by 9 percentage points
over the same period, to 37.7% in 2007). These
positive developments have led to an increase
in participation rates and have compensated
for the negative impact on labour supply of the
Prepared by R. Gomez-Salvador.12
See Annex 2 for a description of this dataset. The need to 13
achieve international comparability means that the LFS dataset
uses standardised and widely accepted defi nitions of e.g.
employment and unemployment, as adopted by ILO. These
constitute the basis of the Eurostat LFS. It should be noted that
these defi nitions differ from those adopted by countries in their
national defi nitions of labour market status, where international
comparability is not necessary.
Suggesting, for example, underemployment-that is, the lower 14
hours worked per year-in the euro area may not be voluntary
(or may not be a matter of choice). See Leiner-Killinger,
Madaschi and Ward-Warmedinger (2005) for a discussion of
institutional arrangements reducing hours of work.
16ECB
Occasional Paper No 87
June 2008
slowdown in population growth rates. However,
projections of future labour supply show that
under current policies, the labour force will start
declining soon due to a fall in the size of the
working age population.
3.1 POPULATION AND THE LABOUR FORCE 15
In 2007, the euro area labour force (the
employed plus unemployed) included over
148 million people, out of a total population
of 318 million and a working age population
(ages 15 to 64) of around 209 million. This
implies a participation rate (labour force divided
by working age population) of close to 71%.
Of the labour force, the number of employed
persons reached around 137 million in 2007,
leading to an employment rate (employment
divided by working age population) of 65.5%,
and the number of unemployed persons
was around 11 million, translating into an
unemployment rate (unemployment divided
by labour force) of 7.5%. Both euro area
participation and employment rates were below
those recorded in the United States (75.3%
and 71.8% respectively in the United States in
2007), while the euro area unemployment rate
was higher (4.7% in the United States) – see
last column of Table 1.
Table 1 also summarises recent and past
developments in population and labour market
indicators. Developments are presented in two
ways: trend developments, to control (to the
extent possible) for cyclical effects; and recent
developments, which divide the past decade
into two fi ve-year periods (of relatively higher
and lower economic growth – see average real
Prepared by R. Gomez-Salvador.15
Table 1 Population, working age population, participation, labour force, employment and unemployment in the euro area and the United States
(average year-on-year growth rates (%), unless otherwise indicated)
Trend developments Recent developments Level 1)
1984-1995 1996-2007 1996-2001 2002-2007 1983 2007
Euro area Population 0.3 0.4 0.3 0.6 291.9 318.8
Working age population 0.5 0.4 0.3 0.5 188.5 209.3
Participation rate 2) 0.2 0.5 0.4 0.6 63.3% 70.8%
Labour force 0.8 1.1 0.9 1.3 119.3 148.3
Population effect 3) 0.5 0.4 0.3 0.5
Participation rate effect 3) 0.3 0.7 0.6 0.8
Employment 0.6 1.6 1.6 1.4 107.2 137.1
Total hours n.a. 1.3 1.1 1.3 n.a. 222,400
Unemployment 1.9 -2.1 -4.2 0.1 12.1 11.1
Employment rate 2) 0.1 0.6 0.7 0.6 56.9% 65.5%
Unemployment rate 2) 0.1 -0.3 -0.6 -0.1 10.1% 7.5%
Real GDP 2.8 2.5 2.8 1.9
US Population 1.1 1.1 1.2 1.0 234.3 302.6
Working age population 1.1 1.4 1.4 1.3 148.3 195.6
Participation rate 2) 0.3 -0.1 0.0 -0.3 73.2% 75.3%
Labour force 1.5 1.2 1.4 1.0 108.5 147.3
Population effect 3) 1.1 1.4 1.4 1.3
Participation rate effect 3) 0.4 -0.2 0.0 -0.3
Employment 2.0 1.3 1.6 1.0 97.9 140.4
Total hours n.a. 1.2 1.3 1.0 n.a. 267,427
Unemployment -2.6 -0.4 -1.3 0.5 10.6 6.9
Employment rate 2) 0.5 -0.1 0.1 -0.2 66.0% 71.8%
Unemployment rate 2) -0.3 -0.1 -0.1 0.0 9.8% 4.7%
Real GDP 4.0 3.7 3.9 2.8
Sources: Eurostat, BLS and ECB calculations. Note: Euro area data refer to the second quarter of each year, while US data are annual averages. 1) In millions, unless otherwise indicated. 2) Average year-on-year changes (percentage point). 3) Contributions to average year-on-year growth rates (percentage points).
17ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY
GDP in Table 1). In the past decade, overall
labour market developments have been quite
favourable for the euro area as a whole. As
a result, the positive gap in participation
and employment rates between the United
States and the euro area has narrowed and
is now closer to those prevailing in the early
1970s (see Chart 1). Positive developments
refl ect a particularly strong increase in female
participation, while male participation rates
actually fell below levels prevailing in 1983.
These developments took place in a context
of broadly stable growth of the euro area’s
working age population.
Labour force developments can be
decomposed into two effects. First, the
population effect, i.e. changes in the working
age population for given participation rates
and, second, the participation rate effect,
i.e. changes in the participation rate for a
given working age population. Comparing
average annual growth rates in the period
1996 to 2007 with those in the period 1984
to 1995, the increase in the participation
rate effect (0.4 percentage point – from
0.3 percentage point to 0.7 percentage point)
more than compensated for the slight trend
decline in the growth rate of the working
age population (the population effect
-0.1 percentage point – from 0.5 percentage
point to 0.4 percentage point), allowing the
growth rate of the labour force to increase
(from 0.8% to 1.1%).
Looking at the same developments within the
past decade shows two interesting results. First,
the participation rate contribution has increased
over time, something that contrasts with the
expected pro-cyclicality of participation rates,
Chart 1 Overall participation and employment rates for the euro area and the United States
(percentages)
Euro area
United States
Participation rate Employment rate
60
65
70
75
80
60
65
70
75
80
1970 1976 1982 1988 1994 2000 200655
60
65
70
75
55
60
65
70
75
1970 1976 1982 1988 1994 2000 2006
Participation rate males Participation rate females
60
70
80
90
60
70
80
90
1983 1986 1989 1992 1995 1998 2001 2004 200645
55
65
75
45
55
65
75
1983 1986 1989 1992 1995 1998 2001 2004 2006
Sources: European Commission, Eurostat, BLS and ECB calculations. Notes: 15 to 64 year olds. See Chart 11 in Appendix 3 for employment rates of males and females.
18ECB
Occasional Paper No 87
June 2008
as periods of relatively low economic growth
normally tend to discourage workers from
participating in the labour force (as discussed
in Chapter 2).16 This suggests that labour
market developments have recently been related
to factors independent of the cycle, such as
changes in the composition of the labour force
and cohort effects. Second, the contribution
from population growth also increased, in
contrast with the trend decline in population
growth rates observed since the early 1980s.
3.2 PARTICIPATION RATES 17
3.2.1 PARTICIPATION RATES BY INDIVIDUAL
CHARACTERISTICS IN THE EURO AREA
The participation rate, and its evolution over
time, is not the same for all groups inside the
working age population. Characteristics such as
gender, age, qualifi cation level and origin have
a strong impact on the observed rate of labour
market participation. The EU-LFS provides
detailed information on the socioeconomic
characteristics of the working, unemployed and
inactive populations in the euro area. Data on
gender and age are available, and the survey
provides information on level of education,
by distinguishing between low (completion of
lower secondary education or less), medium
(completion of up to a diploma of upper
secondary education) and high (holding a
diploma of tertiary education) qualifi cation
level.18 Information on work experience,
training or on-the-job qualifi cations is not
available. Furthermore, information on
country of origin is limited to the nationality
of the survey respondent 19 and distinguishes
The cyclical behaviour of participation in the euro area is 16
documented in Genre and Gomez-Salvador (2002).
Prepared by J. De Mulder.17
The precise defi nition of the educational levels is provided in 18
Annex 2.
Nationality and origin do not necessarily provide the same 19
information, as immigrants or their descendants can have
obtained the nationality of the considered country.
Table 2 Euro area participation rates according to different subdivisions
(employment rates in brackets)
Average annual change (percentage points) Level (%) Trend developments Recent developments
1984-1995 1996-2007 1996-2001 2002-2007 2007
Total 0.2 (0.1) 0.5 (0.6) 0.4 (0.7) 0.6 (0.6) 70.8 (65.5)
Excluding the effect of changes in the population composition 1) n.a. (n.a.) 0.3 (n.a.) 0.1 (n.a.) 0.4 (n.a.)
According to gender
Males -0.3 (-0.4) 0.2 (0.3) 0.1 (0.4) 0.2 (0.2) 78.4 (73.2)
Females 0.6 (0.5) 0.7 (0.9) 0.6 (0.9) 0.9 (0.9) 63.2 (57.8)
According to age
15-24 years old -0.7 (-0.5) 0.0 (0.3) 0.0 (0.6) 0.0 (0.0) 44.0 (37.3)
25-54 years old 0.5 (0.2) 0.4 (0.6) 0.3 (0.7) 0.5 (0.5) 84.6 (79.1)
55-64 years old -0.3 (-0.4) 0.9 (0.9) 0.2 (0.3) 1.5 (1.5) 46.4 (43.4)
According to education level 2)
Low n.a. (n.a.) 0.4 (0.6) 0.3 (0.7) 0.5 (0.5) 63.5 (57.6)
Medium n.a. (n.a.) 0.2 (0.4) 0.1 (0.3) 0.4 (0.4) 80.4 (75.3)
High n.a. (n.a.) 0.0 (0.2) -0.1 (0.3) 0.2 (0.2) 88.3 (84.7)
According to nationality 3)
Nationals n.a. (n.a.) 0.3 (0.5) 0.4 (0.8) 0.2 (0.3) 70.9 (65.9)
Other EU15-citizens n.a. (n.a.) 0.2 (0.3) 0.0 (0.5) 0.4 (0.2) 73.7 (67.6)
Non EU15-citizens n.a. (n.a.) 0.6 (0.8) 0.2 (0.5) 1.1 (1.1) 69.6 (59.3)
of 12 new EU member states n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) 77.1 (68.7)
of non EU27-countries n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) 68.3 (57.7)
Sources: EU-LFS (spring data), ECB and NBB calculations. Note: 15 to 64 years old, except for the subdivision according to education level (25-64 years old). EU15 refers to those countries that were EU Members prior to 2004. The 12 new EU Member States include the countries joining the EU since 2004. 1) Calculated by weighting the participation rates of 18 subgroups of the population of working age (subdivided according to gender, age and education level) with the structure of the population of working age in 1995. 2) EU-LFS data concerning education level only available from 1992 onwards. 3) EU-LFS data concerning nationality only available from 1995 onwards.
19ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLYbetween nationals (persons having the
nationality of the considered country), other
EU citizens and citizens of other (non-
EU) countries. Although level of education
and nationality are not perfect measures
of qualifi cation and country of origin,
they provide the best available proxies for
constructing a long time series to investigate
developments in labour supply according to
these characteristics.
Consideration of the participation rates of these
different groups shows that, on average, female,
low-skilled, young and older persons and non-
EU citizens participate less in the labour market
than other groups (see Table 2). However, with
the exception of the young, the increase in the
participation rate of these groups has accelerated
over the last decade, with the result that they have
at least partially caught-up to the participation
level of other groups. Over the period 1996 to
2007, the average increase in participation has
been largest for females (0.7 percentage point),
older workers (0.9 percentage point) and non-EU
citizens (0.6 percentage point). In contrast, the
participation rate of 15-24 year olds stabilised
during this period.
Participation and employment rates are
still highest for males, prime-aged workers
(25-54 year olds), the highly educated and
EU citizens. Participation is substantially
higher for males than for females in all age
groups (see Table 27, Annex 3). Participation
for both genders is highest between the ages
25 to 49, but while some 90% or more of males
in this age group participated in the labour
market in 2007, this was only the case for
about 77% of females. Female participation
is strongly affected by family status. Until
the age of 49, female participation rates are
clearly lower when a woman has a partner and
when there are dependent children. Similar
differences between genders are also found for
employment rates.
The lower participation rate for younger
persons is often linked to their pursuit of
education. If this results in individuals
obtaining a diploma of upper secondary or, in
particular, tertiary education, the immediate
downward effect on participation rates of
studying longer is compensated afterwards
by higher labour market participation (and
employment) as the education level rises.
Rates of labour market participation, and in
particular employment, remain highest for
the highly educated, but the participation
rate gap related to education is gradually
getting smaller. This catch-up is attributable
to females. Their participation increased for
all three education levels, but the increase
was stronger for those with the lowest level
of education.
Nevertheless, the positive impact of higher
education on labour market participation is still
much stronger for females. In 2007, moving
from a low to medium education level increased
the female participation rate by 24 percentage
points, and from medium to highly skilled by
another 9 percentage points. For males, the
increases were 9 and 5 percentage points
respectively. The remaining gap between female
and male participation is therefore mainly due
to the low skilled. In 2007, 78% of 25-64 year
old low-skilled males participated in the labour
market, 20 while only 50% of their female
counterparts did the same.
Turning to the breakdown by nationality, in
2007, differences in labour market participation
across nationality groups within countries were
relatively small; the highest participation (and
employment) rate is found for citizens of the
12 new member states of the EU. The apparently
almost equal participation of nationals and non-
nationals nevertheless hides two important
facts. First, compared with the corresponding
fi gures for EU citizens, the labour market
participation of non-EU nationals is especially
low for women and middle-aged persons.
Perhaps surprisingly, this is also the case for
highly skilled non-EU citizens, whose
participation rate is comparable to that of
The participation behaviour of older people is treated in Box 6.20
20ECB
Occasional Paper No 87
June 2008
medium-skilled EU-citizens, and particularly
for highly skilled females.21 In relative terms,
this group seems to experience substantially
more problems entering the labour market than
do less skilled immigrants.22 Second, although
non-EU-citizens do not appear to participate in
the labour market much less than nationals on
average, they are less likely to be employed (in
2007 by 8.2 percentage points), and are thus
signifi cantly more often unemployed (discussed
further in Chapter 5).
3.2.2 PARTICIPATION RATES IN THE EURO AREA
COUNTRIES
Labour market participation rates also vary across
euro area countries, ranging in 2007 from 78.5%
in the Netherlands to 62.5% in Italy (see Table 3).
At about 76% to 77%, participation rates were
also relatively high in Germany and Finland,
while Greece, Luxembourg and Belgium were
among countries with the lowest participation
rates at 67% or less. In comparison, employment
rates ranged from 76% of the working age
population in the Netherlands to 59% in Italy. By
means of comparison, in Denmark, Sweden and
the UK, levels of participation and employment
rates are still much higher than in the euro area
and only comparable with the levels in the
Netherlands and Finland.
Looking at trend developments, participation
rates have increased in all euro area countries
over the last decade, and to a greater extent than
in the preceding decade (for almost all countries
for which data are available for the 1980s 23). The
In the absence of more detailed data on immigration, it is not 21
possible to provide details on the explanation of this fi nding.
However, one might speculate that family reunifi cation and
integration issues and/or regulations governing access to the
labour market of non-EU workers may play a role.
A similar observation can be made for ethnic Germans 22
immigrating into Germany from former socialist countries. On
average, the unemployment rate of highly skilled ethnic German
immigrants (with the exception of engineers) is higher than that
of their medium-skilled counterparts. This can be explained,
at least partly, by a lack of applicability of existing skills. For
example, the human capital of lawyers depreciated immediately
and almost completely upon arrival.
No trend for the years 1984 to 1995 could be calculated for 23
Austria and Finland (where the EU-LFS started only in 1995) or
for Slovenia (fi rst EU-LFS collected in 1996).
Table 3 Overall participation rates in euro area countries
(employment rates in brackets)
Average annual change (percentage points) Level (%)Trend developments Recent developments
1984-1995 1) 1996-2007 2) 1996-2001 3) 2002-2007 4) 2007 5)
Belgium 0.2 (0.3) 0.4 (0.4) 0.3 (0.6) 0.5 (0.3) 66.7 (61.6)
Germany 0.3 (0.3) 0.4 (0.4) 0.1 (0.2) 0.7 (0.6) 75.6 (69.1)
Ireland 0.0 (0.1) 0.9 (1.2) 1.0 (1.8) 0.8 (0.6) 72.2 (68.9)
Greece 0.0 (0.0) 0.6 (0.6) 0.5 (0.3) 0.6 (0.8) 67.0 (61.5)
Spain 0.5 (0.3) 0.9 (1.6) 0.6 (1.8) 1.2 (1.4) 71.5 (65.8)
France -0.1 (-0.3) 0.2 (0.3) 0.2 (0.5) 0.2 (0.1) 69.6 (63.6)
Italy -0.1 (-0.3) 0.4 (0.7) 0.4 (0.6) 0.4 (0.7) 62.5 (58.9)
Luxembourg 0.0 (0.0) 0.4 (0.4) 0.6 (0.7) 0.2 (0.0) 65.6 (63.0)
Netherlands 0.9 (1.1) 0.8 (1.0) 1.1 (1.6) 0.5 (0.3) 78.5 (76.0)
Austria n.a. (n.a.) 0.2 (0.2) -0.1 (-0.1) 0.5 (0.4) 73.7 (70.3)
Portugal 0.0 (0.1) 0.5 (0.4) 0.7 (1.1) 0.3 (-0.2) 73.7 (67.6)
Slovenia n.a. (n.a.) 0.5 (0.6) 0.2 (0.4) 0.7 (0.8) 71.7 (68.3)
Finland n.a. (n.a.) 0.4 (1.0) 0.8 (1.6) 0.0 (0.4) 77.3 (71.3)
Euro area 0.2 (0.1) 0.5 (0.6) 0.4 (0.7) 0.6 (0.6) 70.8 (65.5)
Denmark 0.1 (0.3) 0.1 (0.3) -0.1 (0.3) 0.2 (0.2) 80.3 (77.3)
Sweden n.a. (n.a.) 0.2 (0.3) 0.1 (0.6) 0.3 (0.0) 79.9 (74.3)
United Kingdom 0.3 (0.4) 0.0 (0.2) 0.0 (0.5) 0.0 (0.0) 75.0 (71.1)
United States 0.3 (0.5) -0.1 (0.0) 0.0 (0.1) -0.3 (-0.2) 75.5 (72.0)
Sources: EU-LFS (spring data), ECB and NBB calculations.Note: 15 to 64 years old1) 1987-1995 for Spain and Portugal.2) 1997-2007 for Slovenia and 1996-2006 for the United States.3) 1997-2001 for Slovenia.4) 2002-2006 for the United States.5) 2006 for the United States.
21ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLYNetherlands outperformed all other euro area
countries, moving from the lowest participation
rate in 1983 to the highest rate in 2007. During
the past decade, the increase in participation
was strongest in Spain and Ireland. In other
countries, like Luxembourg, Italy and Belgium,
which also had low participation levels in 1996,
the advance was more limited.
However, over the last decade, the apparent
acceleration of participation rates for the
euro area as a whole is not widespread across
individual countries. In six countries (Belgium,
Germany, Greece, Spain, Austria and Slovenia)
the increase in participation was clearly higher
in 2002-07 than between 1996 and 2001.
However, participation growth rates more or less
stabilised in France and Italy and decelerated in
the remaining fi ve countries.
Although remaining lower than for males,
female participation is especially high in Finland
and the Netherlands, but low in Italy, Greece
and Luxembourg (see Annex 3, Table 28).
Low-skilled persons participate substantially
more often in Portugal, and only to a relatively
limited degree in Belgium and Italy. Together
with Spain, Portugal also registers the highest
participation rates of non-EU-citizens. The latter
group participates only to a rather limited degree
in Belgium and the Netherlands.
3.2.3 COMPOSITIONAL EFFECTS IN PARTICIPATION
Although over time, there has been a broad-
based increase in participation across the various
sub-groups considered, the observed evolution
of the euro area aggregate participation rate
can in part be explained by the effects of
changes in the composition of the working
age population.24 Ceteris paribus, this factor
appears to explain almost half of the observed
increase in participation over the last decade 25
(see Table 2), mainly as a result of the gradual
rise in average education level and the shift in
the age structure of the population (the share
of young people in the working age population
decreased in favour of the 25-54 year olds). But
during the most recent fi ve-year period, when
the participation rate increased on average by
0.6 percentage point per year, the impact of
changes in composition fell to one-fourth. Thus
the larger contribution (0.4 percentage point)
to the increase in labour market participation
over the last fi ve years is attributable to a real
underlying increase in participation behaviour.
For individual countries, over the last decade
the contribution of the population structure
effect to participation 26 was especially important
in Italy and Greece. For these countries, the
population effect explained more than three
quarters of the observed participation rate
increase. In Belgium and Ireland it accounted
for about half of the increase, and in Austria,
Finland and France its impact was limited to
one-fi fth or less. In these latter three countries,
the positive impact of the rise in education level
was partly offset by a negative impact from the
ageing of the population towards the 55-64 age
group. During the 2002-07 period, the latter
phenomenon became more widespread,
implying that the population effect was in
general more limited across euro area countries.
In addition to infl uencing the development of
the overall participation rates, differences in the
population structure also have an impact on the
observed participation rate level across euro
area countries.27 As can be seen in Chart 2 for
the year 2007, the impact of the population
As participation structurally depends on factors like gender, age 24
and education level, a shift in the relative shares of these groups
in the population affects the development of the observed overall
participation rate.
For this aim, detailed EU-LFS data on the participation rates 25
of 18 subgroups of the population of working age – subdivided
according to gender (men, women), age (15-24, 25-54, 55-64)
and education level (low, medium, high) – were weighted by
using the population composition of 1995. As (the evolution
of) participation also differs substantially according to
nationality, ideally this factor should also be taken into account.
Unfortunately, for the 1990s, EU-LFS data for several euro area
countries do not provide this subdivision. This makes such a
decomposition exercise unreliable.
Due to a lack of data, no reliable results are available for 26
Germany, Spain, Luxembourg, the Netherlands and Slovenia, or
for Denmark, Sweden and the United Kingdom.
Indeed, for two countries differing only in population structure, 27
the country with, for instance, proportionally more national
highly skilled men aged 25 to 54 will have a higher observed
overall participation rate than the country where the population
proportionally consists of more young or more older, low-skilled
non-EU women.
22ECB
Occasional Paper No 87
June 2008
structure is particularly important for some
countries.28 The effect is most pronounced for
Slovenia, where the observed participation rate
is close to the euro area average, but, having a
population with almost no foreigners and
relatively few low-skilled persons, its rate would
be clearly below this average if it were to have
the same population composition as the euro
area as a whole. Also in countries such as
Luxembourg, the Netherlands and Finland, the
positive effect of population composition is
important, mainly due to a relatively high-
skilled population.29 Conversely, the
participation rates in Italy and Portugal are
considerably higher if one corrects for the
population structure effect, as, according to the
EU-LFS, the population of both countries is on
average less skilled than in the euro area as a
whole. Excluding this population structure
effect, the range of participation rates among
euro area countries is reduced, but still remains
large. The Netherlands still has the highest
participation rate among euro area countries,
while the lowest adjusted rates are found in Italy
and Luxembourg.
3.2.4 COHORT EFFECTS IN PARTICIPATION 30
This section considers to what extent trend
developments in the euro area participation
rate can also be attributed to so-called cohort
effects. These effects are due to differences in
labour market participation across birth cohorts
emerging as a result of different individual
participation choices made early in life
(e.g. regarding fertility, maternity leave and/
or education) and persisting throughout the
life-cycle. Beyond potential crowding-out
effects stemming from the size of the cohort
entering the labour market and differences
in human capital investment over time, these
cohort effects are likely to refl ect evolving
preferences, social norms and/or institutions.31
Chart 3 shows participation rates by age group
of different cohorts in the euro area population.
A cohort refers to persons in the same age
group, thus persons born within a particular
Due to detailed data availability concerning nationality in the 28
EU-LFS for 2007, it was possible to do this decomposition exercise
using 54 population groups, obtained using the combination
of the following factors: gender (men, women), age (15-24,
25-54, 55-64), education level (low, medium, high) and nationality
(nationals, other EU-citizens, non-EU-citizens). To exclude the
population structure effect, the country specifi c participation rates
of these groups were weighted by using the population structure
of the euro area as a whole. For Spain, Ireland, Denmark, Sweden
and the United Kingdom, the results suffer from reliability
problems, as too large a proportion of the population was not
subdivided according to the four factors used.
For Luxembourg, this exercise is less conclusive, as currently 29
38% of persons working in the country are commuters. They
reside in one of the neighbouring countries, and are therefore
part of the population of those countries. As the EU-LFS data
for Luxembourg consider only the Luxembourg population, they
do not refl ect the situation of the whole labour force.
Prepared by A. Balleer and J. Turunen.30
So-called cohort effects generally encompass any factor 31
associated with a particular birth year, e.g. general economic
conditions or crowding-out effects. Empirical evidence suggests
that the size of the cohort entering the labour market greatly
affects participation. There are many explanations for this,
e.g. depressed earnings (see Welch (1979), Berger (1985) and
Korenman and Neumark (1997)), also formulated as the “relative
income” hypothesis, i.e. large cohorts experience lower incomes
relative to their expectations (Easterlin 1980). Finally, large
events like WWII may result in cohort effects as Acemoglu et
al. (2002) fi nd evidence that, owing to men going to war, more
women worked and also stayed in the labour market in the US.
Chart 2 Overall participation rate in 2007: correction for the population structure effect 1)
(differences, in percentage points, with respect to the euro area average)
-10
-8
-6
-4
-2
0
2
4
6
8
10
-10
-8
-6
-4
-2
0
2
4
6
8
10
observed
adjusted 2)
1 Belgium
2 Germany
3 Ireland
4 Greece
5 Spain
6 France
7 Italy
8 Luxembourg
9 Netherlands
10 Austria
11 Portugal
12 Slovenia
13 Finland
14 Denmark
15 Sweden
16 United Kingdom
17 United States
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Sources: EU-LFS (spring data) and NBB calculations.Note: 15 to 64 year olds.1) For Spain, Ireland, Denmark, Sweden and the UK the adjusted participation rate is not displayed, as the mathematical impact originating from the fact that the population used excludes the EU-LFS respondents for which not all characteristics concerning gender, age, education level and nationality are available, was too large. For the United States, no detailed data available to calculate an adjusted participation rate.2) Remaining difference, after correction for data non-availability and population structure (composition of the population by gender, age, education level and nationality).
23ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY
time period, and is represented by a separate
line in the graph.32 The cohort effect is
measured as the vertical distance in
participation rates between the different cohorts
for a given age. The Charts suggest a substantial
cohort effect for females, but no visible cohort
effects for males. Female participation has
risen by close to 20 percentage points when
comparing the participation rate of the
1943-1947-born cohort with the 1963-1967-
born cohort at the age of 40-44 years old.33 In
addition to the substantial level effect, the
shape of the profi le between ages 20 and 35
changes for the younger cohorts as the kink
that is visible in the profi le of those born
1963-1967 at the ages 30-34 disappears. The
timing of this effect suggests that the
differences across cohorts may refl ect a number
of factors relating to, e.g., improved
possibilities for reconciling family and
employment, and shifts towards postponed
motherhood or longer education. Potentially
due to the latter effect, participation of persons
aged 15-24 has declined between the cohort
born in 1968-1972 and the two youngest
cohorts for both males and females.
Table 4 presents the vertical distance
(i.e. the difference in participation levels)
between the youngest and the oldest cohort
for females between the ages of 25 and 44 in
individual countries over the recent decade.
The increases in levels of female participation
between the youngest and older cohort are
particularly large in Ireland, Greece, Spain,
Luxembourg and the Netherlands, but relatively
small in Austria and France. (For comparison,
the changes are even smaller in the UK and
It is generated by starting in the year 2005 for a particular 32
age group, and tracking the participation rate of each cohort
backwards in time over fi ve-year intervals. It has to be noted
that this procedure does not identify a pure cohort effect, but
rather its interaction with age (a so-called age-cohort). Owing to
the relatively short time series dimension, cohort profi les only
partially overlap.
This is the vertical distance between the participation rate of the 33
group born between 1943 and 1947 and the group born between
1963 and 1967 in Chart 3.
Chart 3 Age-cohort profiles in the euro area
(1983-2007)
born 1943-1947
born 1948-1952
born 1953-1957
born 1958-1962
born 1963-1967
born 1968-1972
born 1973-1977
born 1978-1982
y-axis: participation rate
x-axis: age groups
Females Males
10
20
30
40
50
60
70
80
90
10
20
30
40
50
60
70
80
90
1
15-19
20-24
25-29
1
2
3
30-34
35-39
40-44
4
5
6 55-59
45-49
50-54
7
8
9
2 3 4 5 6 7 8 910
20
30
40
50
60
70
80
90
100
110
10
20
30
40
50
60
70
80
90
100
110
15-19
20-24
25-29
1
2
3
30-34
35-39
40-44
4
5
6 55-59
45-49
50-54
7
8
9
1 2 3 4 5 6 7 8 9
Sources: LFS and ECB calculations.
24ECB
Occasional Paper No 87
June 2008
Sweden and even negative in Denmark.) It has
to be noted that for the ages discussed above,
but also for the very young workers, the shape
of the participation profi les, as well as the levels
reached in 2007, are very heterogeneous across
countries. This may be linked, among many
other reasons, to differences in investment
in education or educational systems across
countries.
The cohort profi les shown in Chart 3 are
potentially infl uenced by both business cycle
factors and evolving institutions (which may
also help to explain the country differences).34 35
Overall, a continued increase in the proportion
of women in the labour force and demographic
changes which shift the relative share of the
labour force between birth cohorts imply that
cohort effects will continue to affect euro area
labour supply in the future. Box 1 considers
future developments in labour supply further.
See e.g. Genre et al. (2007) for a study on the role of institutions 34
for participation rates by age and gender. In this study, lagged
participation tries to capture cohort effects for older women
(those between 55-64). This term is statistically signifi cant.
Estimating a cohort-based model of participation for the euro 35
area and most euro area countries using the model presented in
Aaronson et al. (2006) and Fallick and Pingle (2007) which controls
for business cycle and age effects, confi rms that cohort effects are
statistically signifi cant for females and are robust to period effects as
measured by an indicator of the business cycle, while cohort effects
for males are not statistically signifi cant for the euro area. The
estimation suggests a negative effect on participation for the female
cohorts born between 1922 and 1940 (entering the labour market
around the early 1940s and early 1960s) and a positive effect for
those born between 1966 and 1973 (entering in the late 1980s and
early 1990s). The higher participation-propensity of females born
in the late 1960s and early 1970s has therefore contributed to the
increase in female participation in the euro area. The overall pattern
of increasing cohort effects for women appears similar to that
observed in the United States, see Figure 8 in Fallick and Pingle
(2007). In addition, the results suggest that the cycle has strongly
infl uenced the participation rates of the youngest cohorts.
Table 4 Participation differences for females between cohorts in single countries. Change in participation from 1997 to 2007
(percentage points and levels in 2007)
Country Age group 25-29 30-34 35-39 40-44
1997-2007 2007 1997-2007 2007 1997-2007 2007 1997-2007 2007
Belgium 7.4 84.7 8.2 80.7 17.7 81.1 24.5 80.2
Germany 7.2 78.2 13.0 80.8 14.4 81.5 16.7 83.8
Ireland 19.6 82.4 29.2 75.1 34.7 70.3 34.6 67.9
Greece 24.4 77.8 19.2 73.4 21.7 74.4 23.4 71.6
Spain 21.7 80.6 29.9 78.1 36.3 73.8 26.3 72.8
France 5.6 81.2 8.1 80.2 10.8 82.8 12.2 84.2
Italy 2.8 62.7 11.4 68.9 11.7 67.1 16.9 64.7
Luxembourg 9.9 75.1 25.6 80.4 23.1 72.3 25.6 68.0
Netherlands 20.5 86.1 30.5 84.7 25.1 81.6 29.6 81.9
Austria -3.3 76.7 1.8 78.0 6.9 83.4 7.4 85.9
Portugal 12.5 84.9 13.9 87.9 19.8 87.6 14.5 84.7
Slovenia 3.3 79.0 -1.2 81.1 -1.8 85.2 1.1 89.7
Finland 9.6 77.4 14.2 77.9 16.6 77.7 20.8 78.1
Denmark -5.1 82.4 -5.2 84.3 -1.4 86.8 -3.5 85.4
Sweden 2.6 83.5 4.9 87.8 1.5 88.4 -0.1 88.8
United Kingdom 10.8 76.5 10.1 74.9 5.8 76.5 4.1 79.3
United States 1) 0.2 74.9 -0.7 74.1 -2.2 74.0 -1.0 77.0
Sources: LFS and ECB calculations. 1) Developments for the United States are 1995-2007, information is taken from Aaronson et al. (2006) for 1995 and from the Bureau of Labour Statistics for 2007.
25ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLYBox 1
PROJECTIONS OF LABOUR SUPPLY IN THE EURO AREA AND EURO AREA COUNTRIES 1
Demographic developments
A key determinant of labour force developments
is the underlying demographic trend. In
particular, the evolution of birth rates gives an
early indication of future population structure.
Chart A shows that birth rates in the euro area
tended to follow a downward trend from the
mid 1960s until the mid 1990s, remaining
broadly stable thereafter. This general trend
holds for most euro area countries, with
considerable heterogeneity in fertility rate
levels across countries (as indicated by the
distance between the maximum and minimum
birth rates). Ireland and France emerge as the
countries with the highest birth rates in recent
years, and Germany, Italy and the Netherlands
are among those with the lowest. The observed
thirty-year decline in birth rates should lead to
a prolonged decline in the share of prime-aged workers in the total population in the near future.
Since this group exhibits the highest participation rate in the labour market (see Section 3.2.1),
this should also translate into a slowdown in the growth of the labour force.
Future participation rates and labour force developments
On the basis of population projections and assumptions about group-specifi c participation rates,
it is possible to make projections of labour force developments in the euro area.2
According to the latest Eurostat population projections, the working age population is expected
to grow slightly up to 2011 and then decline (although at varying intensity) across the rest of
the forecast horizon (see Table). Under the assumption that participation rates by gender and
age groups remain constant at the 2007 level, the euro area labour force would shrink by an
average of 0.5% per year over the period 2007-2050. This refl ects the continued ageing of
the working age population and, therefore, the previously mentioned decline in the share of
prime-aged workers. Indeed, the overall participation rate would fall by 1.5 percentage points
to 69.4% in 2050 (see Table). As a result of population ageing, the old-age dependency ratio
1 Prepared by R. Gomez-Salvador and A. Novo.
2 Population projections are based on Eurostat’s “baseline variant” for the working age population on January 1st of each year, based on
assumptions on fertility, life expectancy and net migration. In particular, the projections are based on the assumption that the share of
the net migration over the total population remains stable at 0.4% over the forecast horizon. The labour force developments are then
obtained following the methodology described in Shimer (1998).
Chart A Crude birth rates in the euro area 1), 2)
5
10
15
20
25
5
10
15
20
25
2005
average
max
min
1960 1965 1970 1975 1980 1985 1990 1995 2000
Sources: Eurostat and ECB calculations. 1) Ratio of the number of births during the year, per 1,000 inhabitants. 2) Average, maximum and minimum birth rates across euro area countries.
26ECB
Occasional Paper No 87
June 2008
(i.e. the elderly population divided by the working age population) is expected to increase from
around 25% in 2007 to around 55% in 2050.3, 4
For euro area countries, the impact of population structure on the future overall participation rate
is negative in almost all cases, but to varying degrees. Together with the general decline in the
working age population, this will lead to a decline in the future labour force in most countries.5
An alterative scenario
Based on a European Commission ageing report,6 a second scenario for the euro area can be
considered. Relative to the baseline scenario, which assumes that group-specifi c participation
rates remain stable at their 2007 level, the alternative scenario projects higher participation
rates after 2007, particularly among women, whose participation rates have been increasing
over recent decades, and older workers, due to recently enacted public pension system reforms
(see also Section 4.1.3 and Box 6). Under this scenario, the euro area labour force would contract
by 0.3% per year on average over the period 2007-2050, that is, by less than in the previous
scenario. Indeed, the overall participation rate in this alternative scenario would increase by
4.2 percentage points to 75.1% in 2050, to around 5 percentage points higher than in the scenario
presented above. However, the positive contribution from increased labour market participation
would only result in positive developments in the labour force until 2015 (the labour force
growing by 0.4% per year on average). From that year onwards, the negative contribution coming
3 It is worth mentioning that two periods can be distinguished with regard to the impact of ageing on the participation rate. From 2007 to
2030 the participation rate is projected to fall by around 2.5 p.p to 68.5%, and then, as the share of prime-aged workers starts to increase
again, to recover by around 1 p.p in 2050 (see Table). However, the positive contribution coming from this increased participation in
the last part of the projection horizon is not expected to have any signifi cant impact on labour force developments, being outweighed by
the continued decline in the working age population (see Chart B).
4 For detailed country-specifi c and euro area charts on the projected developments in dependency ratios, see Maddaloni, et al (2006).
5 The only two exceptions are Ireland and Luxembourg, where positive developments in the working age population may more than
offset the decline in participation, and therefore translate into an increase in the labour force (see Table).
6 European Commission (2006, 2007). This second scenario does not assume a general rise in education levels, but analyses the effects
of expected demographic and labour market developments given the present enrolment and cost situation.
Working age population, participation rate and labour force developments 1)
Working age population 2) Participation rate 3) Labour force 2)
2007-10 2011-30 2031-50 2007-50 2007 2010 2030 2050 2007-10 2011-30 2031-50 2007-50
Belgium 0.3 -0.3 -0.2 -0.2 66.7 66.1 65.2 65.3 0.0 -0.4 -0.2 -0.3
Germany -0.1 -0.6 -0.7 -0.6 75.6 75.9 74.4 74.5 -0.1 -0.7 -0.7 -0.6
Ireland 0.8 0.6 -0.2 0.3 72.2 72.0 70.3 70.8 0.9 0.5 -0.2 0.2
Greece 0.2 -0.3 -1.0 -0.6 67.0 67.0 63.6 64.8 0.2 -0.6 -0.9 -0.6
Spain 0.2 -0.2 -1.2 -0.6 71.5 71.3 67.0 68.7 0.2 -0.6 -1.1 -0.7
France 0.3 -0.2 -0.2 -0.1 69.8 69.2 68.3 68.7 0.1 -0.2 -0.1 -0.2
Italy -0.1 -0.5 -1.0 -0.7 62.5 62.3 58.6 60.3 -0.2 -0.8 -0.9 -0.8
Luxembourg 1.0 0.5 0.5 0.5 65.6 64.6 63.2 63.2 0.6 0.4 0.5 0.4
Netherlands 0.3 -0.2 -0.1 -0.1 78.5 78.2 77.4 78.0 0.1 -0.2 -0.1 -0.1
Austria 0.3 -0.3 -0.5 -0.4 74.9 74.6 72.0 72.2 0.2 -0.5 -0.5 -0.4
Portugal 0.0 -0.4 -0.9 -0.6 73.7 73.8 71.1 72.0 0.0 -0.5 -0.9 -0.6
Slovenia 0.0 -0.6 -0.8 -0.6 71.7 71.8 69.0 69.4 -0.1 -0.8 -0.7 -0.7
Finland 0.2 -0.5 -0.2 -0.3 77.3 76.8 77.5 77.2 0.0 -0.5 -0.3 -0.3
Euro area 0.1 -0.4 -0.7 -0.5 70.8 70.8 68.5 69.4 0.0 -0.5 -0.6 -0.5
Sources: Eurostat and ECB calculations. 1) Working age population derived from Eurostat projections (baseline scenario). Overall participation rate derived by keeping participation rates by gender and age group constant at the 2007 level. 2) Annual growth rates. 3) Levels expressed as percentages.
27ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY
3.3 HOURS WORKED 36
To evaluate labour utilisation, it is important to
look at not only the (relative) number of people
participating in the labour market (the extensive
margin), but also the number of hours worked per
employed person (intensive margin). Indeed, in
the context of population ageing and shrinkage
of the total labour force, both channels can be
used to increase the effective labour supply.
To obtain a fuller picture of labour supply, this
section considers average yearly hours of work
at an aggregate level from 1991 onwards using
data from the OECD.37 Unfortunately these data
are not available for different sub-groups of the
labour force, so this analysis uses EU-LFS data
on usually-worked weekly hours.38 The data on
both concepts are comparable, since the OECD
uses the EU-LFS data as input for calculating
annual hours.39 In addition, data on annual
and weekly working hours are closely linked
(see Chart 12 in Annex 3).
3.3.1 HOURS WORKED IN THE EURO AREA AND IN
THE UNITED STATES
The main difference between hours of work in
the United States and Europe is the number of
hours worked in annual terms. According to the
most recent OECD data, an average euro area
worker worked 1,672 hours in 2005,40
Prepared by J. De Mulder and R. Gomez-Salvador.36
See Annex 2 for a short description of this dataset.37
The EU-LFS also provides information on actual weekly 38
working hours, but this concept is – more than the usual working
hours – infl uenced by one-off factors (for instance exceptional
absences or overtime work) during the reference week of the
survey.
The OECD fi gures on hours worked per year are obtained by 39
combining EU-LFS information on weekly hours (usual hours,
overtime work and hours on additional jobs) with the number of
weeks worked per year.
As Slovenia is not an OECD member, there are no OECD data 40
available for the euro area as a whole. Therefore a euro area
average of the other twelve countries was calculated, weighted
by employment (number of persons at work). Before 1995,
no data are available for Austria and Finland, so the euro area
fi gures for the period 1991-1994 correspond to the weighted
average of ten countries.
from demographic developments would dominate, and labour force growth would start declining
by 0.5% per year on average until 2050 (see Chart C).
In sum, even if the most dominant recent trends in labour market participation, such as the
increase in female participation continue, the labour force can be expected to decline in the near
future, due to the projected fall in the working age population.
Chart B Euro area labour force projections – constant participation rates 1)
Chart C Euro area labour force projections – varying participation rates 2)
labour force (left-hand scale)participation rate (right-hand scale)
100,000
110,000
120,000
130,000
140,000
150,000
160,000
170,000 76
62
64
66
68
70
72
74
1983 1995 2007 2016 2025 2037100,000
110,000
120,000
130,000
140,000
150,000
160,000
170,000
1983 1995 2007 2019 2031 204362
64
66
68
70
72
74
76
labour force (left-hand scale)
participation rate (right-hand scale)
Sources: Eurostat and ECB calculations. 1) Constant participation rates across gender and age groups (15-24, 25-29, 30-34, 35-39, 40-44, 45-49, 50-54, 55-59 and 60-64) at the 2007 level. 2) Varying overall participation rates at the country level.
28ECB
Occasional Paper No 87
June 2008
considerably less than the 1,922 hours worked in
the United States. This suggests that regulations
on hours of work and annual holidays may play
an important role in explaining the differences
between annual and weekly hours in the United
States and euro area. Over the past decade,
average annual hours of work also fell faster in
the euro area than in the United States (by
respectively 0.3% and 0.1% per worker per year).
However, since 2003, a minor increase in annual
hours has been observed in the euro area (see
Chart 4).41 With regard to weekly hours, EU-LFS
data show that the decline in average working
time per worker had already started in the early
1980s, and the reduction in hours of work in
Europe in recent years has come mainly from a
relative decline in hours per week. The average
working week was 40.1 hours in the euro area in
1983, decreasing to 37.4 hours per week in 2007.
Comparable data for the United States are only
available for the period 1996-2005. The data
show that the average working week in the United
States was around 39 hours in 1996, comparable
to hours worked in the euro area. However, since
then the length of the average working week has
roughly stabilised in the United States.42 In 2005,
the latest available year, it was 38.5 hours.
The information on annual hours worked per
worker can be combined and compared with
data on the number of people employed to allow
a qualifi cation of the developments in total hours
worked in the euro area (see Chart 5). This shows
that following a slight decline, both the number
of people employed and total hours worked show
a similar upward trend since the mid 1990s,
although employment increased faster than
total hours worked (while total hours appears to
have been more pro-cyclical).43 Moreover, the
picture of employment developments in the last
decade changes somewhat when labour input is
measured in hours (rather than in the number of
people). It appears that instead of the decline
in average growth rates between 1996-2001
and 2002-07 recorded by the employment rate
(as shown in Table 1), the growth in total hours
has remained broadly stable (at 1.1%), similar to
the growth rate of total hours in the Unites States.
Finally, it is worth mentioning that this different
pattern has an impact on measured productivity
growth in the euro area, which shows a less
marked slowdown in the second half of the
This is driven by the smaller number of annual weeks worked 41
in Europe, fewer overtime hours worked on the main job in
Europe, and fewer hours on additional jobs in Europe.
According to the Groningen Growth and Development Centre 42
(GGDC), hours worked were also stable in the US over the
whole of the 1980s and 1990s.
Indeed, looking at the period 2001-2004, it appears that hours 43
worked per worker have acted more as a buffer to economic
conditions, given that total hours declined slightly and
then recovered, in line with the economic slowdown, while
employment growth remained slightly positive.
Chart 4 Hours worked per worker in the euro area and the United States
(euro area (EA) vs. United States (US))
2,000
1,900
1,800
1,700
1,600
1,5001983 1987 1991 1995 1999 2003 2007
46
44
42
40
38
36
annual hours EA (left-hand scale)
annual hours US (left-hand scale)
weekly hours EA (right-hand scale)
weekly hours US (right-hand scale)
Sources: Eurostat, OECD and ECB calculations.
Chart 5 Developments in employment, total hours worked, and productivity per person and per hour in the euro area
88
92
96
100
104
108
112
1991 1993 1995 1997 1999 2001 2003 2005-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
total hours - 1991=100 (left-hand scale)
employment - 1991=100 (left-hand scale)
productivity per hour (right-hand scale)
productivity per person (right-hand scale)
Sources: OECD, EU-LFS and own calculations. Employment in thousands and Total hours in millions.
29ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY1990s when measured in hours worked instead
of number of employed persons.
3.3.2 HOURS WORKED IN THE EURO AREA BY
DIFFERENT WORKER GROUPS 44
This section examines whether the number
of hours worked per worker in the euro area
varies by worker group. In addition to the
characteristics of gender, age, education level
and nationality, professional status (employee
or self-employed) and work regime (part-time
or full-time work) may also affect average hours
across countries.45
An important explanation for the downward
trend in euro area hours worked per worker per
week is the substantial increase in part-time
work.46 As can be seen in panel B of Table 5,
the average part-time worker worked 20 hours
per week in 2007, compared with 41.5 hours
for a full-time worker. In 1983, only 9% of
the working age population was working part-
time; however, by 2007, this fi gure had risen to
19%. This development is linked to the rise in
female participation described in the previous
section (see also Section 4.2). In 2007, 35%
of all working women had a part-time job,
compared with only 7% of men; thus, the greater
importance of part-time work for women also
explains their lower average number of hours
worked.
At an average of 34 hours per week in 2007,
young persons worked the fewest hours. This
fi gure mainly refl ects the fact that many young
people work part-time as they combine their
studies with a job.47 There is little difference
Prepared by J. De Mulder. 44
The latter two factors were not treated in the section on 45
participation rates, as they only concern the working population,
whereas the labour force is composed of both the working and
the unemployed.
In the EU-LFS, the full-time versus part-time distinction is based 46
on the declaration by the respondent, except in the Netherlands,
where this subdivision is made according to whether the
respondent usually works at least 35 hours (full-time) or less
(part-time).
In 2006 (last year for which this information is available), an 47
average working student aged 15 to 24 worked for roughly
13 hours, which has a considerable downward infl uence on the
average number of hours worked by young persons.
Table 5 Hours worked per worker in the euro area
Panel A: Total
Average annual change (%) Level (hours) Trend developments Recent developments
1984-1995 1996-2005 1996-2001 2002-2005 2005
Annual hours n.a. -0.3 -0.3 -0.1 1,672
1984-1995 1996-2007 1996-2001 2002-2007 2007
Weekly hours -0.3 -0.3 -0.3 -0.2 37.4
Excluding the effect of changes in the population composition 1) n.a. -0.2 -0.3 -0.2Panel B: Weekly hours in 2007 according to different subdivisions (hours)According to work regime According to gender
Full-time work 41.5 Males 40.9
Part-time work 20.0 Females 32.9
According to age According to education level15-24 years old 33.7 Low 38.2
25-54 years old 37.8 Medium 37.1
55-64 years old 37.3 High 38.4
According to nationality According to professional statusNationals 37.4 Employees 35.9
Other EU15-citizens 36.9 Other workers 45.5
Non EU15-citizens 36.1
of 12 new EU member states 37.3
of non EU27-countries 35.3
Sources: EU-LFS (spring data), ECB and NBB calculations. Note: 15 to 64 years old, except for the subdivision according to education level (25-64 years old). 1) Calculated by weighting the hours worked by 36 subgroups of the population of working age (subdivided according to gender, age, education level and professional status) with the structure of the population of working age in 1995.
30ECB
Occasional Paper No 87
June 2008
in hours worked per week between prime-aged
and older workers, according to education level
and nationality. In contrast, there is signifi cant
variation in the average hours worked by
professional status. In 2007, employees worked
on average 36 hours per week, while the average
week was 45.5 hours long for other workers
(mainly self-employed).
The decline of weekly working time since 1983
was broadly based and comparable for most
worker groups. Nevertheless, the decrease was
somewhat stronger for females, due to the faster
up-take of part-time work, and for the younger
and older generations. The latter developments
most likely refl ect the larger proportion of
young people pursuing tertiary education and
the higher participation rate of older workers,
who remain at work longer in life but reduce
their average weekly working time.
3.3.3 HOURS WORKED PER WEEK IN THE EURO
AREA COUNTRIES
As with labour market participation rates,
signifi cant variation in the average yearly hours
of work is apparent across euro area countries
(see Table 6).48 In 2005, average working time
ranged from almost 2,000 hours per worker per
year in Greece, to some 1,400 hours in the
Netherlands.49 In most other countries, the
average person worked between 1,600 and
1,800 hours (compared with 1,600 hours in
Denmark, Sweden and the United Kingdom).
During the past decade (until 2005), developments
As mentioned before, no data for yearly hours worked are 48
available for Slovenia.
These observed differences are largely attributable to differences in 49
the composition of employment, mainly as regards the occurrence
of part-time work and self-employment. Indeed, if only full-time
employees are considered, average working hours in Greece only
exceed those in the Netherlands by some 100 hours.
Table 6 Hours worked in euro area countries
Annual hours 1) Weekly hours Average annual change (%) Level
(hours) Average annual change (%) Level
(hours) Trend Recent developments Trend developments Recent developments
1996-2005 1996-2001 2002-2005 2005 1984-1995 1996-2007 1996-2001 2002-2007 2007
Belgium 0.0 0.0 0.1 1,681 -0.5 -0.1 -0.1 -0.2 37.1
Germany -0.4 -0.4 -0.6 1,622 -0.5 -0.5 -0.4 -0.6 35.6
Ireland -0.9 -1.3 -0.2 1,729 -0.5 -0.9 -1.3 -0.6 36.3
Greece -0.1 -0.1 0.0 1,995 -0.4 -0.2 -0.1 -0.3 42.5
Spain 0.0 -0.1 0.0 1,791 -0.3 -0.2 -0.2 -0.3 39.3
France -0.6 -0.6 -0.6 1,592 -0.4 -0.1 -0.7 0.5 38.0
Italy 0.1 -0.2 0.5 1,730 -0.1 -0.2 -0.2 -0.2 38.5
Luxembourg -0.7 -0.6 -0.9 1,637 -0.2 -0.6 -0.5 -0.7 36.7
Netherlands -0.2 -0.3 0.0 1,417 -1.1 -0.5 -0.7 -0.4 30.9
Austria 0.2 0.1 0.3 1,729 n.a. 0.0 -0.2 0.3 39.0
Portugal -0.8 -1.3 -0.1 1,775 -0.5 -0.6 -0.9 -0.2 39.6
Slovenia n.a. n.a. n.a. n.a. n.a. -0.4 -0.1 -0.6 40.3
Finland -0.1 0.0 -0.4 1,682 n.a. -0.1 0.1 -0.2 37.9
Euro area -0.3 -0.3 -0.1 1,672 -0.3 -0.3 -0.3 -0.2 37.4
Denmark 0.3 0.8 -0.3 1,586 -0.3 0.0 0.3 -0.3 35.7
Sweden 0.5 0.9 -0.2 1,558 n.a. 0.1 0.2 -0.1 36.7
United Kingdom -0.4 -0.2 -0.7 1,662 0.0 -0.3 -0.3 -0.3 37.2
United States 2) -0.1 -0.2 0.0 1,922 n.a. -0.1 -0.1 -0.1 38.5
Sources: EU-LFS (spring data) and OECD calculations. Note: 15 to 64 years old. 1) OECD annual hours data not available for the 1980s, so the trend development for 1984-1995 could not be calculated. 2) For annual and weekly hours in the United States, the considered trend period and the fi rst recent development period start in 1997 instead of 1996. For weekly hours, the second recent development period concerns 2002-05, and the level in the last column concerns 2005.
31ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLYhave diverged substantially across countries.
While average working hours rose in Austria,
considerable decreases were recorded in Ireland,
Portugal, Luxembourg and France recorded
considerable decreases. Since 2002, the
downward trend has continued in the latter two
countries. In other countries, especially Italy,
average annual working hours actually increased.
The order of the euro area countries according to
the average number of weekly hours worked is
roughly comparable to the one based on average
annual hours. In 2007 average weekly working
time was highest in Greece (42.5 hours), and
by far lowest in the Netherlands (30.9 hours).
In Denmark, Sweden and the United Kingdom,
the situation was comparable to the euro area
average, with 36 to 37 hours.
Average weekly hours worked fell in almost all
European countries over all considered periods;
only in Austria and France was an increase
observed during the last fi ve-year period. The
decrease was strongest in the Netherlands
(especially in the 1984-1995 period) and in
Ireland (during the last decade). An important
factor in both the level and downward trend in
average weekly hours worked is the rate of
part-time work, which may be linked to both
institutional features and individual preferences.
For example, Greece, which has the highest
average working week, also experiences one of
the lowest rates of part-time work among the
euro area countries. Similarly, the Netherlands,
with the shortest working week, has the highest
rate of part-time work. Furthermore, the
decrease in the average hours worked per week
in the Netherlands coincided with a strong
increase in part-time work following changes to
part-time work legislation in 1982.50
3.3.4 COMPOSITIONAL EFFECTS IN HOURS
WORKED
The overall impact of the evolution of the
composition of employment on observed
working hours 51 is rather limited for the euro
area as a whole (see panel A of Table 5). Overall,
the downward infl uence of the higher proportion
of females and employees in employment was
partly compensated by a decreasing share of
young workers.
Contrary to the euro area average, changes in
the composition of the employed population
seem to have had a considerable impact in some
countries.52 In Belgium, Spain, Finland, Greece
and Italy, they accounted for almost the entire
development of average working hours over
the last decade, implying that the underlying
working hours of the various subgroups
have hardly changed. In France, a downward
development took place, which was partly
counteracted by changes in the population
structure. In most other countries, the effects of
changes in the population structure have been
limited over the last fi ve years.
Part of the observed differences in average hours
worked between countries in 2007 can also be
attributed to the different country-composition
of employment.53 For example, the short
working week in the Netherlands can partly be
explained by a relatively larger share of young
workers in employment than on average in the
euro area, while part of the longer working week
in Greece can be attributed to a relatively high
proportion of male workers and self-employed.
For a discussion of the explanations for the differences in hours 50
of work across countries, see also Leiner-Killinger, Madaschi
and Ward-Warmedinger (2005).
For this ceteris paribus exercise, detailed EU-LFS data 51
concerning the hours worked of 36 subgroups of the working
age population – subdivided according to gender (men, women),
age (15-24, 25-54, 55-64), education level (low, medium, high)
and professional status (employee or self-employed) – were
weighted using the employment composition of 1995. In the
previous section, these four factors were identifi ed as the most
important distinguishing characteristics concerning hours
worked in the euro area. The work regime was not taken into
account, since defi nitions of part-time and full-time work depend
on the national legal system (for instance, a person working
38 hours per week is working (more than) full-time in France
and Belgium, but is working part-time in other countries). In
addition, this variable is closely linked to gender.
The size of the population effect depends on the relative size 52
of the different evolutions inside the employed population. In
most countries, the share of females and employees is growing,
having a downward effect on average hours worked. On the
other hand, the proportion of young persons in employment is
decreasing, pushing upwards the average working time.
Again, detailed EU-LFS data, subdivided according to gender, 53
age, education level and professional status, were used. To exclude
the population structure effect, the country-specifi c working hours
of these groups were weighted using the employment structure of
the euro area as a whole.
32ECB
Occasional Paper No 87
June 2008
After adjusting for the structure of employment,
the spread of weekly hours worked inside the
euro area decreases by one fourth, to 9 hours per
week. Nevertheless, working weeks still appear
to be longest in Greece and – by far – shortest in
the Netherlands (see Chart 6).54
3.4 IMMIGRATION 55
Both the quantity and the quality of labour
supply in the euro area are important in order to
maximise welfare and future potential growth.
Immigration provides one channel for the euro
area to increase its labour supply along both
dimensions. Furthermore, it may help alleviate
shortages of particular skills in the labour
market, improving the allocation of labour
resources. It may offset some of the negative
effects of demographic change and, since
immigrants tend to be more mobile than native
workers, it may also help the labour supply
adjust to economic shocks. Immigration to the
euro area has increased signifi cantly over recent
decades. Chart 7 presents the net fl ow of
migrants into the euro area and into the United
States since 1980. It shows that net migration to
the euro area has been higher than to the United
States since 1997.56 Furthermore, migration to
the euro area prior to the late 1990s was mainly
the result of guest-worker programmes (1950s
and 1960s), family-reunifi cation (1970s) and
asylum-seeking (late 1980s and early 1990s,
when political events and ethnic confl icts
increased). More recently, euro area immigration
appears to have entered a new phase, possibly
as a result of EU enlargement, globalisation, and
changing immigration policies in some
countries, with an increase in the number of
immigrants (also of different origin) looking for
employment.
In the case of Greece, the relatively long hours even after this 54
adjustment may refl ect in part the relatively high share of retail
and tourist businesses with long operating hours.
Prepared by M. Ward-Warmedinger.55
This measure therefore estimates the immigration of individuals 56
from outside each area, it does not consider cross-state or cross-
euro area migration.
Chart 7 Net migration to the euro area and the United States
(in millions of persons)
-0.5
0.0
0.5
1.0
1.5
2.0
-0.5
0.0
0.5
1.0
1.5
2.0
1980 1985 1990 1995 2000 2005
euro areaUS
Source: Eurostat. Note: For this chart only, net migration is the difference between immigration into and emigration from the respective area during the year, often measured as the difference between the total population on 1 January and 31 December for a given calendar year, minus the difference between births and deaths (or natural increase). To the extent that this data does not capture illegal immigration, it may underestimate immigrant fl ows.
Chart 6 Weekly working hours in 2007: corrected for the population structure effect 1)
(differences, in hours, with respect to the euro area average)
-8
-6
-4
-2
0
2
4
6
-8
-6
-4
-2
0
2
4
6
1 Belgium
2 Germany
3 Greece
4 Great Britain
5 Spain
6 France
7 Italy
8 Luxembourg
9 Netherlands
10 Austria
11 Portugal
12 Slovenia
13 Finland
14 Denmark
15 Sweden
16 United Kingdom
17 United State
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
observed
adjusted2)
Sources: EU-LFS (spring data) and NBB calculations. Note: 15 to 64 years old.1) For the United States, no detailed data are available to calculate an adjusted number of weekly working hours.2) Remaining difference, after correction for the population structure (composition of the population by gender, age, education level and professional status).
33ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLYWhilst the potential for signifi cant economic
gains from immigration exists, realising these
gains depends on the characteristics of
immigrants relative to nationals and immigrants’
successful integration into the labour market.
The future challenges facing the euro area
suggest the need for a great variety of skills. For
example, immigration may help meet the future
demand (see Chapter 5) for services such as
nursing, household care, childcare, health care
and eldercare arising from population ageing,
and for high skilled workers. Furthermore, it is
important that immigrants’ skills and
qualifi cations are effectively utilised.57
Table 7 shows that the increase in immigration
to the euro area since the mid-1990s was
composed largely of females, non-EU15
nationals and prime age individuals (25-54 age
group).58 Whilst nearly half of the existing
population of immigrants in 2007 were low-
skilled persons, the fl ow of new immigrants
since 1996 has mainly been made up of medium
and highly skilled labour. This may partly refl ect
selective immigration policies in place in many
EU Member States, which try to attract highly
skilled immigrants (discussed further in
Section 4.3). The increase in immigration from
1996 to 2007 was largest in Spain, Greece,
Portugal and Luxembourg (see Table 29 in
Annex 3). The extent to which migration fl ows
have contributed to labour supply also varies
strongly across euro area countries (see Table 8,
Table 29 in Annex 3 and also Box 2). One
particular form of worker immigration – namely
daily cross-border commuting – has increased
three-fold over the past 10 years (see the Box in
Work by the OECD (2006) “Gaining from migration: towards a 57
new mobility scheme” suggests that the over-qualifi cation rate is
two to three times higher for foreign born relative to native born
in some euro area countries.
In the absence of more detailed data on immigration, it is not 58
possible to provide details on the explanation of this trend.
However, one can speculate that family reunifi cation, an increase
in female labour market participation and the regularisation of
illegal work in some countries may have played a role in these
developments.
Table 7 The composition of the working age population by nationality
Average annual change (p.p)
Level (%)
1996-2001 2) 2002-2007 3) 2007
Total working age population 1)
Nationals -0.1 -0.2 91.9
Other EU15 citizens 0.0 -0.1 1.8
Non EU15 citizens 0.1 0.3 6.3
of 12 new
member states n.a. n.a.5) 0.9
of Non EU27 n.a. n.a.6) 5.3
Non-national working age population 4)
According to genderMales -0.4 0.1 -0.3 0.0 50.0 50.1Females 0.4 -0.1 0.3 0.0 50.0 49.9
According to age15-24 years old -0.7 -0.1 -0.2 -0.2 15.9 17.525-54 years old 0.3 0.2 0.5 0.0 73.5 64.355-64 years old 0.3 -0.1 -0.3 0.2 10.6 18.2
According to education levelLow -0.3 -0.2 -0.6 -0.1 48.6 36.6Medium -0.1 -0.2 0.9 0.2 34.4 42.1High 0.5 0.5 0.4 0.2 16.7 21.0
Participation ratesNationals 0.4 0.2 70.9
Other EU15
citizens 0.0 0.4 73.7
Non EU15 citizens 0.2 1.1 69.6
of 12 new
member states n.a. n.a.5) 77.1
of Non EU27 n.a. n.a.6) 68.3
Employment ratesNationals 0.8 0.3 65.9
Other EU15
citizens 0.5 0.2 67.6
Non EU15 citizens 0.5 1.1 59.3
of 12 new
member states n.a. n.a.5) 68.7
of Non EU27 n.a. n.a.6) 57.7
Source: EU-LFS and ECB calculations. Note: 15 to 64 years old. To the extent that this data does not capture illegal immigration, it may underestimate the stocks and fl ows of immigrants.1) The non-national population is separated into non-national EU15 citizens and non-national non-EU15 citizens. For the period 2005-07, this last group is further split into the 12 new member states (which together with the EU15 form the EU27) and non-national non-EU27 citizens. 2) It is important to note that data for Ireland start in 1998, data for Portugal and the Netherlands in 1999, and data for Slovenia in 2002. 3) Irish data only available until 2004, Italian data only available for 2005-07. 4) The numbers in italics show the respective values for the national working age population. 5) The average annual change 2006-07 for total working age population is 0.1, for participation rates 0.6 and for employment rates 1.4. 6) Average annual change 2006-07 for total working age population is 0.2, for participation rates 0.6 and for employment rates 1.3.
34ECB
Occasional Paper No 87
June 2008
Annex 4). Nevertheless, while the employment
level of migrants from other EU countries is
similar to or even higher than that of nationals
for most euro area countries, overall participation
and employment rates for non-EU nationals
lagged behind those of nationals in 2007,
especially in Belgium, Germany, France, the
Netherlands and Finland. There is some
evidence in some countries of decreasing
employment rates for non-nationals from 2002
to 2007. On the other hand, in Spain, Greece,
Italy and Portugal, the employment rate for non-
EU nationals was even higher than for nationals
in 2007 (see Table 8).59
These differences in participation rates may refl ect a country’s 59
history of immigration. For example, countries where
immigration is a new phenomenon may experience a high
proportion of immigrants who are driven by the economic
motivation to enter the labour market and fi nd a job. Countries
with a longer history of immigration may receive immigrants
driven by more varied and potentially less labour-market-driven
motivations, including, for example, family reunifi cation.
Table 8 Immigration in euro area countries
Country % of working age population 2007 Participation rates 2007 Employment rates 2007
NationalsOther EU15
Non-EU15
of which
12 NM
of which Non-
EU27 NationalsOther EU15
Non-EU15
of which
12 NM
of which Non-
EU27 NationalsOther EU15
Non-EU15
of which
12 NM
of which Non-
EU27
Belgium 90.9 5.3 3.8 0.5 3.2 67.0 69.1 55.7 71.1 53.2 62.4 62.4 40.6 62.1 37.1
Germany 89.6 2.8 7.6 1.0 6.6 76.7 76.7 63.8 73.3 62.4 70.6 69.6 51.5 63.7 49.7
Ireland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Spain 87.1 1.5 11.4 2.3 9.1 70.5 70.2 79.7 82.4 79.0 65.3 62.7 70.0 72.9 69.2
Greece 93.8 0.3 6.0 1.0 5.0 66.6 53.6 73.8 71.7 74.2 61.1 48.8 67.9 65.6 68.4
France 94.1 2.1 3.8 0.2 3.6 70.0 72.8 59.3 68.0 58.9 64.2 67.8 44.9 53.6 44.5
Italy 94.2 0.3 5.5 1.0 4.5 61.9 60.0 73.0 76.7 72.2 58.4 57.9 67.4 71.1 66.6
Luxembourg 57.9 37.7 4.4 1.1 3.3 61.7 71.6 64.9 63.4 65.4 59.4 69.3 57.7 61.4 56.4
Netherlands 95.7 1.6 2.7 0.2 2.5 79.1 78.0 57.6 72.3 56.3 76.7 75.8 51.8 67.7 50.4
Austria 88.7 2.2 9.1 1.6 7.6 74.2 77.2 67.8 75.1 66.3 71.3 73.1 59.5 69.6 57.3
Portugal 96.3 0.4 3.3 0.2 3.1 73.4 74.6 82.8 72.9 83.5 67.5 69.3 70.6 67.8 70.8
Slovenia 99.3 n.a. 0.7 n.a. 0.7 71.7 n.a. 65.7 n.a. 65.1 68.4 n.a. 59.7 n.a. 59.0 Finland 98.0 0.4 1.6 0.4 1.2 77.4 85.1 67.8 80.0 63.9 71.6 77.8 53.6 74.4 47.0
Euro area 91.9 1.8 6.3 0.9 5.3 70.9 73.7 69.6 77.1 68.3 65.9 67.6 59.3 68.7 57.7
Denmark 94.6 1.1 4.4 0.2 4.2 81.2 76.6 61.1 76.3 60.3 78.5 73.8 53.9 71.8 52.9
Sweden 95.0 2.1 2.9 0.3 2.6 80.4 75.4 65.1 74.1 63.9 75.1 70.8 52.3 56.0 51.9
United
Kingdom 92.2 1.8 6.0 1.4 4.6 75.2 76.9 71.0 84.7 66.7 71.4 71.5 65.2 79.6 60.7
Sources: Eurostat, LFS and ECB calculations. Notes: 15 to 64 years old. For Ireland: data available for 1998-2004 only; for Italy: for 2005-07 only; for the Netherlands and Portugal: data start in 1999; for Slovenia: data start in 2002; for Sweden: data start in 1997. For Greece, fi gures from the LFS differ signifi cantly from those of the 2001 Population Census. According to the 2001 Population Census non-EU and other EU15 citizens accounted for 7.7% and 0.5% of the working age population in Greece in that year. Numbers in italics are based on fi gures smaller than the Eurostat’s reliability limits.
35ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY
Characteristics of immigrants for selected countries (2007)
Ireland Spain Italy Austria Immigrants Natives Immigrants Natives Immigrants Natives Immigrants Natives
Quantity
% Total Population 6.6 12.9 5.8 11.3
Gender
% Male 50.6 50.2 49.9 50.7 49.0 50.1 49.5 49.8
Age
15-24 20.2 23.5 18.1 16.3 14.7 15.6 17.0 18.0
25-54 73.5 62.1 76.9 66.6 81.1 65.4 72.0 64.7
55-64 6.3 14.4 5.1 17.0 4.2 19.0 11.0 17.3
Education
Low 20.8 39.1 46.3 50.3 51.8 48.9 38.4 23.9
Medium 28.5 36.7 33.8 21.2 37.4 39.0 45.7 61.9
High 38.3 23.3 19.1 27.0 10.9 12.1 15.8 14.2
Participation Rate 66.6 68.8 78.5 70.5 72.4 61.9 69.6 74.2
Unemployment Rate 6.3 4.4 12.0 7.3 7.6 5.7 10.8 3.9
Source: EU-LFS.Note: 15-64 year old.* Data for Ireland refer to 2004.
Box 2
RECENT EXPERIENCES WITH IMMIGRATION IN EURO AREA COUNTRIES 1
Although most European countries have experienced an increase in immigration during the
last 20 years, country experiences with immigration have differed (in terms of the magnitude of
infl ows, the characteristics of immigrants and the impact of immigration on the native population).
While some countries have a long history of immigration (e.g. Germany, France and Austria),
there are several countries for which large-scale immigration is a relatively new phenomenon (e.g.
Ireland, Spain or Italy). Numerous factors account for these cross-country differences, including
historical ties with a host country, a common language, geographical proximity and the extent of
labour market opportunities, which can affect both the magnitude and duration of immigration to
a particular country. This box focuses on the immigration experiences of Spain, Italy, Ireland and
Austria, which have attracted the majority of permanent immigrants in recent years.
Immigration in Spain and Italy
The immigration experiences of Spain and Italy have been fairly similar in recent years. In both
countries immigrants have become an important and a fast growing share of the population.
Spain has faced the highest net migration rates in Europe, with an average of 700,000 new
immigrants per annum during the last two years. This consistently high infl ow pushed up the
percentage of non-nationals in the Spanish population from 2% in 2000 to 13% seven years later
(see Table). Italy also experienced consistently high infl ows of foreign workers, although at a
relatively lower rate, increasing the share of non-nationals from 2.7% at the end of 2002 to 6% of
the total population in 2007.
1 Prepared by A. Rosolia, A. Lacuesta, Y. McCarthy and A. Stiglbauer.
36ECB
Occasional Paper No 87
June 2008
Geographical proximity and historical ties are the main factors underlying migration to Spain and
Italy. Immigrants from South America account for over one third of the immigrant population
in Spain, and immigrants from Africa, especially from Morocco, accounted for 20% of the total
stock in 2006. Albania, Morocco and Romania are the main source countries of migration to Italy,
accounting for one-third of all immigrants in 2006. Despite the retention of some restrictions on
the free movement of workers from the 2004 EU new Member States, the fl ow and the share of
immigrants coming from these countries have signifi cantly increased. For example, in Spain,
those origin countries represented 2% of all immigrants in 2000, compared with more than 15%
just seven years later.
Regarding individuals’ characteristics, in both countries, non-nationals are much younger than
nationals. While about 65% of natives are concentrated in the 25-54 age group, this number
increases to above 76% for the non-national population. There are no important educational
attainment disparities between non-nationals and nationals in either country.
Despite the magnitude of immigrant fl ows and the similarity of educational attainment of
immigrants and natives, there is no evidence that migration has reduced the job opportunities
for residents in either country. Indeed, the simple correlation across regions between the
participation and employment rates of Italian citizens aged 25-54 and the share of foreigners in
the same population age-group is not statistically signifi cant.2 Research conducted at the Bank
of Italy relates natives’ labour market outcomes to the presence of foreign citizens in a province
over the decade 1993-2003, taking into account a host of individual characteristics as well as the
characteristics of local labour markets that could affect simultaneously Italian citizens’ labour
market outcomes and the overall presence of foreigners. The results show that male and female
employment and participation display a positive correlation with the presence of foreigners of
the same sex. Carrasco et al (2008) undertake an analysis in this regard for the Spanish case,
fi nding that foreigners have no signifi cant negative impact on natives’ occupational opportunities.
Rather, the increase in foreign workers has occurred at the same time as the increase in female
participation, and some research has linked the two phenomena. This evidence is confi rmed by
independent work of The Economic Bureau of the President in Spain, which provides evidence
that immigration facilitates female participation in the labour market by increasing the supply of
domestic help, thus easing the combination of work and family life for women.
One possible explanation for the lack of a negative relationship between immigrants’ and natives’
labour market outcomes is the fact that highly skilled immigrants have been willing to take jobs
requiring less skill than their educational attainment. Fernández and Ortega (2006) show that
this matching problem is more important for immigrants than for natives (see also Chapter 5).
Moreover, immigrants’ contractual conditions have been more fl exible than those of natives with
the same characteristics. In Spain, for instance, the share of immigrants on a temporary contract
is around 60%, compared with 30% for natives. Unemployment has also been much higher for
non-nationals than for nationals (respectively 12%/8% for non-nationals and 7%/6% for nationals
in Spain/Italy). However, at least part of this difference can be attributed to immigrants’ lack
of experience in their destination country, and over time the gap with the native population is
expected to close (Amuedo-Dorantes and de la Rica, 2007; Fernández and Ortega, 2006).
2 Because immigrants tend to locate where job opportunities are richer, such simple correlations might hide some crowding-out of
comparable native workers.
37ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLYImmigration in Austria
At 11% in 2007, Austria has a large non-national population. At the end of the 1960s, the
immigrant population share was only 2%. Thereafter, it increased gradually to 4% at the end
of the 1980s, then, in the wake of the fall of the “iron curtain” and the Yugoslav wars, a large
number of immigrants entered Austria within just a couple of years. By 1995, the immigrant share
had risen to more than 8%. Since then, it has risen continuously, although at a slower pace. Most
immigrants stay permanently, but seasonal work is also signifi cant. The share of immigrants in
the workforce is more than 12%, with the largest groups of non-national workers coming from
the former Yugoslavia, Germany, Turkey, Hungary, the Czech Republic and Slovakia, Poland
and Romania.
Non-nationals in Austria are on average younger than nationals: 11% of the non-national
population is between ages 55 and 64, compared with 17% of natives. Non-nationals tend to
be less educated than nationals, with 38% of the foreign population holding only a primary
education, compared with 24% of nationals. This partly underlies the fact that immigrants work
disproportionally in industry, in the construction sector, in tourism and agriculture. Seasonal
work and commuting are very common in the latter sectors. Illegal infl ows of immigrant workers
are most likely substantial, particularly in the household and personal service sector, which
includes cleaning and nursing services.
The available empirical evidence suggests that the aggregate effects of immigration on native
workers’ unemployment and wage growth are small or insignifi cant. Econometric studies are
mostly from the mid-1990s and exploit the then-sudden increase of foreign workers. They fi nd
that increased immigration had no negative employment or wage effects for native women. For
almost all groups of men, these studies fi nd only a slight deterioration of employment prospects.
Low-income men also faced lower wage growth, whereas high-income males appeared to
experience a wage gain associated with the increase in immigration.
Austria faces several challenges with regard to the integration of immigrants. For example, the
unemployment rate for non-nationals in 2007 stood at about 11% (whereas that of nationals was
merely 4% − see also Chapter 5). The difference between the performance of immigrant children
and nationals in the PISA studies was also one of the largest observed over all participating
countries (OECD, 2007b). Without appropriate countermeasures, Austria risks perpetuating the
lower labour market chances of immigrants and their descendants.
Immigration in Ireland
Ireland’s experiences with migration changed dramatically during the 1990s as Ireland rapidly
moved from being a country of net emigration to one of signifi cant positive net infl ows. Since
1996, net migration to Ireland has been positive, and the most recent numbers show a net
migration of 71,800 individuals for the twelve months to April 2006 (1.7% of the population).
As a result, there has been a rise in the non-national proportion of the population, from about
6% in 2002 to about 10% in 2006. The change in Ireland’s migration experience has also had
an important impact on labour supply growth. In the period 2000 to 2005, labour supply grew
by almost 3%, with migration accounting for just under half of this growth. By comparison, in
the early 1990s migration subtracted from labour supply growth. This change in the quantity
of immigrants in Ireland was accompanied by an important change in the origin country of
38ECB
Occasional Paper No 87
June 2008
3.5 THE SUPPLY OF KNOWLEDGE AND SKILLS 60
Labour quality can be improved through
investment in human capital. Human capital
consists of the ability, skills and knowledge
embodied in the general population that are
accumulated through schooling, training and
experience.61 This section begins with an
examination of a traditional measure of human
capital − that is, educational attainment − as
captured by the highest level of education
attained by individuals in the euro area. It then
examines alternative measures to account for
the quality of educational attainment. Box 4
assesses the growth in labour quality in the euro
area over time.
3.5.1 EDUCATIONAL ATTAINMENT
Improving the stock of human capital has
been formally identifi ed through the Lisbon
strategy as a key area of potential growth
within the euro area. The level of educational
attainment within a country serves as one
potential indicator of the stock of such capital.
Table 9 shows educational attainment of euro
area countries as captured by the proportion of
the adult population that has received various
levels of education over time.
The average proportion of the 30-54 year
old population with a high level of education
(tertiary education) in the euro area was 16.1%
in 1992. By 1999 this had increased to 20.1%,
and the fi gure stood at 24.6% in 2007. On the
other hand, there has been a fall in the proportion
of the population with only a low level of
education, and a subsequent rise in those with
a medium level. However, the rate of increase
in this measure of the stock of human capital
differs across euro area countries. For example,
Ireland registers the largest increase in persons
with a high level of education between 1999
Prepared by P. Cipollone, Y. McCarthy and K. McQuinn. This 60
contribution has benefi ted from the comments of P. Montanaro,
Bank of Italy.
In much of the initial empirical work addressing this issue, 61
human capital has been measured by educational attainment.
Educational attainment, in this regard, is taken to be the number
of years of schooling received by an individual.
immigrant infl ows. In the late 1980s, over half of immigrants entering Ireland came from the
United Kingdom. However, over time this proportion fell signifi cantly, reaching about 17% in
2006. In recent years, immigrants from the new EU Member States have been the most important
contributors to immigrant infl ows in Ireland.
Focussing on characteristics, Labour Force Survey data available to 2004 (see Table) show that
immigrants in Ireland are young relative to the native population. Data for 2004 indicate that
73.5% of the non-national population is in the prime working age category, 25 to 54 years old,
compared with 62.1% of natives. Immigrants in Ireland also tend to be more highly educated than
natives, as over 38% of non-nationals hold tertiary education compared to 23.3% of nationals.
Despite this higher educational attainment, non-nationals face higher unemployment rates relative
to nationals, at 6.3% and 4.4% respectively in 2004. 3 A study conducted by Barrett et al (2006)
using data from a national household survey for 2003 found that immigrants in Ireland tended to
hold occupations of a lower level compared with those held by natives, once characteristics such
as age and education had been taken into account. Barrett and McCarthy (2007) examined the
earnings of immigrants relative to natives in 2004, fi nding that immigrants earned 18% less than
natives on average, controlling for education and years of work experience. This gap was more
pronounced for immigrants from non-English speaking countries and for those immigrants with
a third-level qualifi cation.
3 However, this fi nding of a higher rate of unemployment for immigrants is not unusual, as immigrants could possess lower levels of
location-specifi c human capital when they arrive in a host country. Over time it would be hoped that immigrants would assimilate into
the labour market and that this rate of unemployment would fall.
39ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY
and 2007 at 12.8 percentage points, followed
by Spain with a 9 percentage point increase.
Furthermore, in 2007, the proportion of the
population with a high level of education was
highest in Finland, Belgium and Ireland, while
it was lowest in Italy, Portugal and Austria.
An examination of educational attainment by
gender shows that both males and females in the
euro area registered an increase in educational
attainment between 1992 and 2007, and by 2007
there was very little difference between male
and female educational levels in the euro area.
Table 9 Educational Attainment of 30-54-Year-Old Population
(% of population by highest level of education attained)
1992 1999 2007Country Low Medium High Low Medium High Low Medium High
Belgium 47.1 30.4 22.5 40.4 31.8 27.7 29.4 37.3 33.3
Germany 17.4 58.8 23.8 17.5 57.4 25.1 14.1 60.1 25.8
Ireland 56.9 25.8 17.3 44.1 36.0 19.9 30.6 36.7 32.8
Greece 60.5 26.0 13.5 46.2 35.0 18.8 36.4 40.1 23.5
Spain 76.1 10.9 12.9 62.2 16.4 21.4 46.6 23.0 30.5
France n.a. n.a. n.a. 37.1 42.1 20.8 29.4 43.4 27.2
Italy 64.8 27.2 8.1 53.9 35.4 10.7 45.3 40.8 13.9
Luxembourg 64.1 22.8 13.1 35.0 45.5 19.5 32.7 38.9 28.4
Netherlands n.a. n.a. n.a. 33.8 42.6 23.6 24.5 44.1 31.4
Austria n.a. n.a. n.a. 22.6 62.3 15.1 18.5 63.1 18.5
Portugal 79.3 9.0 11.7 80.9 9.9 9.1 72.2 13.9 13.9
Slovenia n.a. n.a. n.a. 24.3 60.0 15.7 17.1 59.6 23.3
Finland n.a. n.a. n.a. 23.7 42.8 33.5 15.4 44.2 40.4
Euro area 47.4 36.5 16.1 39.3 40.6 20.1 31.9 43.5 24.6
Denmark 23.3 53.7 23.0 19.5 51.8 28.7 23.1 43.2 33.7
Sweden n.a. n.a. n.a. 20.5 49.2 30.3 12.1 55.8 32.0
United Kingdom 49.4 31.0 19.6 36.2 36.0 27.8 26.6 40.7 32.6
Sources: Eurostat, LFS and ECB calculations.Notes: 30 to 54 years old. Data include vocational training to the extent that a qualifi cation has been gained. See Annex 2 for details on this and defi nitions of low, medium and high. See Table 30 in Annex 3 for this information for prime age individuals only. See Table 31 for information on education attainment by cohort.
Table 10 Educational Attainment: Subject specialisation
(2006)
SubjectGeneral
Programmes
TeacherTraining
and Education
HumanitiesLanguages
and Art
Social Science Businessand Law
Science Math
Computing
Engineeringand
Manufacturing Agriculture
Health and
Welfare Services
Belgium 6.6 9.1 6.2 22.2 6.8 28.4 1.6 12.7 6.4
Germany 6.0 5.2 3.8 28.3 2.3 34.2 3.0 8.7 8.4
Ireland 50.8 3.9 4.0 11.1 8.8 8.1 1.7 6.4 5.2
Greece 47.6 3.0 5.9 13.1 4.1 13.7 1.4 6.0 5.2
Spain 30.2 6.1 5.8 22.8 6.1 15.6 1.3 8.6 3.5
France 1.6 0.9 10.2 32.2 9.3 29.3 3.7 8.0 4.8
Italy 0.1 10.2 13.0 31.5 11.8 22.4 2.2 4.8 3.9
Luxembourg 8.3 5.7 9.3 29.7 7.0 15.0 1.8 7.2 16.1
Netherlands 8.3 9.4 4.9 27.2 3.5 19.2 2.9 15.6 8.9
Austria 8.3 4.1 3.8 26.6 1.3 34.6 4.5 5.5 11.2
Portugal 3.3 8.8 20.9 25.7 19.0 12.7 0.9 6.0 2.7
Slovenia 8.5 4.7 2.1 26.3 1.2 37.9 3.2 5.5 10.6
Finland 13.4 3.2 4.4 17.6 2.3 29.3 4.9 12.8 11.9
Euro area 15.0 5.9 7.5 24.7 6.8 22.6 2.4 7.9 7.3
Denmark 1.8 4.8 7.2 30.3 5.0 26.7 3.1 15.1 6.0
Sweden 11.3 8.3 4.9 20.7 2.9 26.6 2.3 15.6 7.5
United
Kingdom 0.7 6.6 11.7 25.5 11.0 21.5 1.7 14.0 7.4
Sources: LFS and ECB calculations. Ireland refers to 2005 data.
40ECB
Occasional Paper No 87
June 2008
For example, in 1992, respectively 19.4/12.8%
of males/females aged 30-54 years had a high
level of education in the euro area. These fi gures
increased to 21.8/18.4% in 1999 and 24.7/24.5%
in 2007.
Table 10 provides information on the type of
skills acquired through education in euro area
countries. More specifi cally, the table shows
educational attainment across euro area countries
in 2006 (the latest year available) according to
the main subject studied. The results show that
“engineering” (manufacturing and construction)
and “social science, business and law” are
generally among the top two most studied
disciplines in the majority of euro area countries.
In Ireland, Greece and Spain, however, “general
programmes” is the most popular area of study.62
In general, there has been little change in the
respective rankings of subject take-up across euro
area countries since 2003, the earliest year for
which data are available.
3.5.2 THE QUALITY OF EDUCATIONAL
ATTAINMENT
While cross-country comparisons of the
quantity of education received are important and
informative, attention has increasingly focussed
more on differences in the quality of education
received across countries. In attempting to
measure the quality of educational attainment, a
growing line of research uses indexes based on
scores obtained by students on cognitive tests.
Cross-country information on such measures
is provided under a number of different
international auspices. The IEA-TIMSS (Trends
in International Mathematics and Science Study)
collects educational achievement data at the
fourth and eighth grades in mathematics and
science. Comparable data are available for 1995,
1999 and 2003.63 ,64 The IEA-PIRLS (Progress
in International Reading Literacy Study)
examines literacy progress at the fourth grade
(9-and 10-year-olds). The fi rst survey was
conducted in 2001 on about 150,000 students
across 35 countries (including fi ve euro area
countries). A second survey was conducted in
2006. The OECD has launched the Programme
for International Student Assessment (PISA),
an internationally standardised assessment of
the competences of 15-year-old students. The
assessment covers the domains of reading,
mathematical and scientifi c literacy, and includes
all euro area countries with the exception of
Slovenia. The fi rst survey, conducted in 2000,
focused mostly on reading literacy; the second
was conducted in 2003 with a major focus on
math; science literacy is the main focus of the
third survey, conducted in 2006. The development
of these databases is of considerable interest,
particularly in the context of studies examining
cross-country income differentials.
“General programmes” refers to all other categories of education. 62
It is important to note that these categories may be somewhat
sensitive to a country’s educational system. For example, many
students at the degree level in Ireland would take a general ARTs
degree, included under the “General Programmes” defi nition.
However, in another country degree programmes may be more
specialised, allowing them to be categorised under “Social
Science, Business and Law” or “Humanities, Languages and
Art”.
The IEA (International Association for the Evaluation of 63
Educational Achievement) is an independent, non-profi t,
international cooperative of national research institutions and
governmental research agencies, established in 1959.
For more on this, along with a summary of results from both 64
PIRLS and TIMSS, see Montanaro (2007).
Box 3
HUMAN CAPITAL AND GROWTH 1
Initial growth studies incorporating the role of human capital such as Barro (1991), Benhabib
and Spiegel (1994), Barro and Sala-i-Martin (1995) and Sala-i-Martin (1997) had focussed
solely on the role played by educational attainment in terms of the number of years of schooling
1 Prepared by P. Cipollone, Y. McCarthy and K. McQuinn. This contribution has benefi ted from the comments of P. Montanaro,
Bank of Italy.
41ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY
Table 11 compares educational attainment of
euro area countries as captured by the traditional
quantity measure, highest level of education
attained, with the PISA results for 2006 − the
latest year for which data are available. While
the quality and quantity of education are two
different concepts, the results in the table
nonetheless demonstrate a positive correlation
between the rankings of countries in terms of
the proportion of the population with a high
level of education and the rankings based on the
PISA profi ciency scores. For example, Finland
registers the highest proportion of its population
with a high level of educational attainment
in 2006, at 38.7%, followed by Belgium, at
32.6%. Finland also ranks the highest across the
three PISA categories, while Belgium and the
Netherlands are typically among the top four
countries when ranked by both the PISA average
scores and the proportion of the population with
a high level of educational attainment. In this
context, Greece would appear to be an outlier.
It registers among the lowest of the euro area
countries in terms of the average scores across
the three PISA categories, while it enjoys an
average level of high educational attainment
relative to the other euro area countries.
A key message emerging from this section is
that euro area countries have generally been
successful in increasing their stock of human
capital over time. However, there is scope for
further improvement, particularly since some
euro area countries still lag far behind the
average human capital level in the euro area,
as well as the level of human capital in other
advanced economies. Moving beyond the
human capital level of the euro area population,
Box 4 examines developments in the quality
of the euro area workforce between 1992
and 2005. This analysis suggests that labour
quality growth is likely to decline over the next
decade. This slowing of labour quality growth
could have important ramifi cations for labour
productivity growth and potential output. In
an attempt to ensure continued labour quality
growth commensurate with past rates, policy
should be directed towards promoting further
increases in educational attainment and on-the-
job training (see also Section 4.4 and Box 8).
in a country. These studies generally found that educational attainment was positively correlated
with the growth rate of GDP per capita across countries. The provision of databases such as
Trends in International Mathematics and Science Study (TIMSS) and Progress in International
Reading Literacy Study (PIRLS), however, enables studies of income differentials to focus on
the role played by enhanced educational quality. For example, TIMSS formed the basis for the
measurement of labour-force quality in studies by Hanushek and Kimko (2000) and in Barro
(2001). Both of these studies used the TIMSS database as an enhanced indicator of educational
attainment in growth regressions. The Hanushek and Kimko (2000) results are particularly
interesting as they emphasize how the explanation of cross-country growth is affected by the
inclusion of quality measures. Their estimates suggest that a one-standard-deviation improvement
in a test performance (equivalent to 47 score points in PISA 2000 mathematics) could increase
the annual average growth rate of a country by 1 percentage point. 2 Recently, Hanushek and
Wößmann (2007) have extended the Hanushek and Kimko results to 50 countries (from 31),
exploiting all the available information gathered by IEA’s and OCSE PISA survey on test score
in math and science. The study supports the Hanushek and Kimko fi nding of a positive link
between educational quality and GDP growth. Additional work by Bosworth and Collins (2003),
Ciccone and Papaioannou (2005), Coulombe et al. (2004) and Coulombe and Tremblay (2006)
all fi nd that improvements in educational quality strongly outweigh increases in educational
quantity in infl uencing economic growth.
2 See Montanaro (2007) for a more detailed discussion of this issue.
42ECB
Occasional Paper No 87
June 2008
Box 4
LABOUR QUALITY GROWTH IN THE EURO AREA AND EURO AREA COUNTRIES 1
Increases in the supply of highly educated and more experienced workers have the potential
to contribute positively to economic growth. Standard measures of labour input, such as total
hours worked, ignore these changes in the composition of the workforce, typically leading to an
underestimation of the contribution of labour to growth (see OECD, 2001). This box presents
estimates of the trend in labour quality growth for the euro area and euro area countries. The
estimates of labour quality are constructed in two steps. In a fi rst step, microdata are used to
derive weights for a number of worker groups with different characteristics. These weights refl ect
differences in productivity (measured by estimated relative wages 2) across workers groups, e.g.
those with university level education or more are on average more productive than those with
primary education and are thus given a larger weight. In a second step, these weights are used
to adjust data on total hours worked by worker-country groups to arrive at an index of labour-
quality-adjusted labour input. Labour quality growth is estimated as the difference between
1 Prepared by J. Turunen.
2 More specifi cally, time-varying weights are derived from predicted wages from cross-section regressions of individuals’ wages on their
human capital characteristics such as education and labour market experience (as proxied by age).
Table 11 Educational Attainment of Population and PISA scores, 2006
(% of total population by highest level of education attained; Pisa score)
Country 2006 PISA 2006 PISA 2006Low
EducationMedium
EducationHigh
EducationProfi ciency in Reading
Profi ciency in Maths
Profi ciency in Science
Percentage 30-54 year old population Average Score Total
Belgium 30.9 36.4 32.6 501 520 510 1531
Germany 15.0 59.3 25.7 495 504 516 1515
Ireland 31.7 37.1 31.2 517 501 508 1526
Greece 37.2 39.3 23.5 460 459 473 1392
Spain 48.3 22.3 29.4 461 480 488 1429
France 31.1 42.9 26.0 488 496 495 1479
Italy 46.1 40.5 13.4 469 462 475 1406
Luxembourg 34.0 41.6 24.4 479 490 486 1455
Netherlands 25.3 43.7 31.0 507 531 525 1563
Austria 18.4 63.0 18.7 490 505 511 1506
Portugal 72.1 14.2 13.7 472 466 474 1412
Slovenia 17.7 60.4 22.0 494 504 519 1517
Finland 15.6 45.8 38.7 547 548 563 1658
Euro area 32.8 43.1 24.1 491 497 503 1491
Denmark 17.1 46.3 36.6 494 513 496 1503
Sweden 12.6 56.3 31.1 507 502 503 1512
United Kingdom 27.2 41.7 31.1 495 495 515 1505
Sources: Educational attainment data are from Eurostat, LFS and ECB calculations. PISA data are from the OECD.
43ECB
Occasional Paper No 87
June 2008
3 MAIN TRENDS IN
LABOUR SUPPLY
quality-adjusted and raw total hours worked. Estimates of labour quality growth are based on a
number of assumptions and data sources and should thus be interpreted with great caution.3
Focussing on the period 1992-2005, estimates in Schwerdt and Turunen (2007) suggest that
euro area labour quality has grown on average by 0.48% year-on-year. Relatively strong labour
quality growth in most of the 1990s, driven by an increase in the share of those with tertiary
education and workers in prime age (35-54 years of age), was followed by lower labour quality
growth towards the end of the 1990s, possibly refl ecting the impact of robust employment
growth resulting in the entry of workers with lower skills. The euro area estimate of labour
quality growth refl ects substantial diversity across individual countries for which reliable
estimates are available, with estimates ranging from the lowest in Finland (0.18%) to the highest
in Spain (0.84%) (see Chart A). In line with other studies (see, e.g. Jorgenson, 2005, for the
G7 countries), the rise in the average level of educational attainment is the main driver of the
increase in labour quality over time, with a consistent relative contribution to labour quality
growth also across euro area countries. Chart A shows that overall labour quality growth has
signifi cantly benefi ted from increasing shares of highly educated employees in countries like
Spain, Ireland and Austria. Other countries, such as Finland and Germany, do not show such
signifi cant increases in the share of highly educated employees and, consequently, have lower
3 In particular, estimates of labour quality growth are based on the key assumption that relative marginal products of worker types are
refl ected in their relative wage rates. Various characteristics of labour markets, such as discrimination, union bargaining, signalling and
mismatch, may result in violations of this assumption. However, due to a lack of more direct measures, wages remain the best available
proxy of worker productivity. Furthermore, individual labour market experience is not directly observable in available household data.
Therefore, as is standard in the literature, age is used to proxy labour market experience. The weights are derived separately for men and
women, allowing e.g. for the age-earnings profi les to differ across gender. Levenson and Zoghi (2007) construct labour quality growth
estimates for the US based on birth-imputed experience measures and age and fi nd that using age results in a slightly lower estimates of
labour quality growth. Schwerdt and Turunen (2007) use detailed information from the European Community Household Panel (ECHP)
and the LFS on total hours worked and wages by worker groups along four dimensions − age (6 groups), education (3), gender (2) and
(for the euro area estimates) country (12) − to construct estimates of labour quality growth. For a more detailed description of the data
and methodology, see Schwerdt and Turunen (2007). Because of the reclassifi cation of education categories in the LFS that occurred in
the late 1990s and other breaks, estimates for country-years in which at least a 5% jump in an underlying share of total hours worked
within a single education category is observed (1998 in Ireland, 1998 in Finland, 1999 in Austria and 2004 in Greece) are excluded
from the calculation of time period averages shown in this box.
Chart A Labour quality growth and the contribution of education
(average annual growth rate in 1992-2005; percent)
Germany
Spain
Belgium
Austria
Ireland
Euro Area
FranceItalyPortugal
GreeceNetherlands
Finland0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
x-axis: growth in labour quality
y-axis: 1st order contribution of education
Source: ECB calculations based on estimates in Schwerdt and Turunen (2007).
Chart B Labour quality growth over time in the euro area
(annual growth rate; percent; 1992 to 2005)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1992 1994 1996 1998 2000 2002 2004 2006
Source: ECB calculations based on estimates in Schwerdt and Turunen (2007).
44ECB
Occasional Paper No 87
June 2008
fi rst-order contributions from education to labour quality growth over this time period. Overall,
these country results are broadly consistent with estimates from other studies.4
Chart B shows the growth in labour quality in the euro area since 1992. Looking forward, owing
to the ageing of the euro area population, the relative share of (the most productive) workers
of prime-age is likely to decline, putting downward pressure on growth in euro area labour
quality in the coming 10-15 years. This effect poses an additional challenge for sustaining labour
productivity growth in the euro area.
4 See Jorgenson (2005) for estimates for Germany, France and Italy; Card and Freeman, (2004) for Germany; Melka and Nayman (2004)
for France; Brandolini and Cipollone (2001) for Italy; Inklaar et al. (2005) for the EU4 (Germany, France, the Netherlands and the
United Kingdom) and EU-KLEMS for a number of European countries (see www.euklems.net).
45ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
4 STRUCTURAL POLICIES AND THEIR EFFECT
ON LABOUR SUPPLY 65
An individual’s labour supply is characterised by
two key decisions: fi rst, whether to participate
in the labour market; and second, once in the
labour market, how many hours to work. Such
decisions are determined by preferences and
budget constraints over the life-cycle and are
greatly affected by institutions and structural
conditions.66 Empirically, a large literature has
documented the signifi cant impact of labour
and product market institutions on labour
supply across countries.67 This literature has
emphasized that the impact of institutions differs
across labour types.68 Institutional changes are
found to have the strongest impact on the labour
supply of individuals with a weaker attachment
to the labour market, such as younger and
older workers, females and (although less
evidence is available) immigrants.69 Against the
background of population ageing, globalization
and technological change, policies which
succeed in attracting, integrating and retaining
a high number of people in productive jobs and
increasing human capital cannot be understated
in their importance for increasing income levels
and for the sustainability of public fi nances in
the Member States.
This chapter thus takes a closer look at
the contributions of structural policies or
institutions to increasing euro area labour
supply. While these institutions will mainly be
considered separately below, the interaction of
different institutions (e.g. active labour market
policies and unemployment benefi ts) and the
dependency of their effect on macroeconomic
conditions is an important issue.70 Particular
consideration is given to institutions affecting
the subgroups in Chapter 3, namely: (i) Tax
and benefi t systems, and the design of pension
and early retirement systems, (ii) “Work/family
reconciliation” policies, including parental
leave, part-time work opportunities and child
care provision, (iii) Policies and institutions
affecting immigration and the integration of
immigrants into the labour market and society,
(iv) Educational and training policies.
This chapter’s main fi ndings include the fact
that reducing disincentives to work, such as
high marginal tax rates, high unemployment
benefi ts, early retirement schemes and weak
work availability requirements, can stimulate the
labour supply and employment of all workers, but
particularly those with a generally more tenuous
attachment to the labour market, such as women
and older workers. Female labour supply in
particular is supported by so called “work/family
reconciliation” policies. These policies help to
reconcile fertility and labour market participation
developments, with positive implications for
long-term labour supply and potential output.
From an economic perspective, the benefi ts
derived from immigration depend on both the
quantity and the characteristics of migrants
entering the euro area. The design of national
policies and institutions is important to set proper
incentives and to facilitate the integration of
immigrants into the labour market and society.
Prepared by A. Balleer, J. Turunen and M. Ward-Warmedinger.65
For example, the participation decision is affected by factors 66
infl uencing the attractiveness of entering the labour market,
relative to remaining outside it, such as the tax system and the
generosity and duration of unemployment benefi ts, institutions
that affect the incentives of workers (and fi rms) to invest in
education and skills, contractual arrangements and regulations
affecting the fl exibility of hours of work including childcare,
parental leave and/or part-time work opportunities. The second
decision is largely determined by the extent to which working
more hours results in higher current or expected net income. In
the case of tight regulation, the key decision may degenerate to a
decision about whether to participate full-time or part-time with
the precise number of hours in these categories fi xed by law or
social partners. This infl exibility may also lead to individuals
deciding not to enter the labour market at all. Additional aspects
of individual labour supply include joint household decisions
and decisions relating to the life-cycle (e.g. how many years
to participate and decisions relating to investment in human
capital).
While there exists a large number of studies that focus on 67
unemployment, employment and participation decisions have
not attracted as much attention in the literature. See for example,
Genre et al. (2005b) on participation rates in the European
Union, and Bassanini and Duval (2006) and Bertola et al.
(2002), who investigate employment and relative employment
of the sub-groups in OECD countries in 1982-2003 and
1960-1996 respectively.
See for example Antunes and Cavalcanti (2007), Genre et al. 68
(2007), Bassanini and Duval (2006), Bertola et al. (2002) and
Nickell and Layard (1999) for more details.
See for example Bertola et al, (2002), Amable et al. (2007), 69
Nickell and Layard (1999).
See Bassanini and Duval (2006), Carone and Salomäki 70
(2001), Bertola et al. (2002), Blanchard and Wolfers (2000),
Killingsworth and Heckman (1986).
46ECB
Occasional Paper No 87
June 2008
Finally, educational systems and the amount
and effi ciency of national resources devoted
to education play a key role in determining
the labour supply of the young and innovation
capacity, overall labour quality within the euro
area and long-term wage levels. It is important
that national education systems are well funded
and effi cient, providing positive incentives for
young people, workers and fi rms to invest in
education and training, and for the effi ciency and
service orientation of educational institutions to
be improved. Furthermore, educational systems
have an important role to play in ensuring a
smooth transition from education to working life
and in providing the labour force with relevant
skills for the future. However, a general caveat
is needed: namely that policies involving higher
government spending must also be considered
from the perspective of the government budget.
The distorting effects of tax increases to fi nance
such policies also need to be taken into account.
4.1 TAX AND BENEFIT SYSTEMS 71
Tax and benefi t systems are a major explanatory
factor of the labour supply developments
observed in the euro area over the last decade.
Depending on their design, tax and benefi t
systems affect individuals’ incentives to engage
in paid employment in several ways. First, they
may affect the decision to enter paid employment.
High income support for persons not in
employment (e.g. unemployment benefi ts, social
assistance, benefi ts from disability schemes,
housing benefi ts) and high taxes on labour, such
as income taxes and social security contributions,
reduce the incentives for moving from inactivity
to activity, from informal (or activities in the
shadow economy) to formal work and from
unemployment to employment. Second, tax and
benefi t systems affect work effort or human capital
formation. In this respect, high labour taxes reduce
incentives to work longer hours, to move from
part-time to full-time work, to increase efforts and
learning to enhance future income prospects, and
to move to jobs with higher productivity. Third,
pension systems affect incentives to stay in the
labour force longer. Against this background, this
subsection describes through what channels the
features of euro area countries’ tax and benefi t
systems, as well as changes therein, may have
supported the rise in labour supply that was
identifi ed in the previous chapter.
4.1.1 UNEMPLOYMENT BENEFIT SYSTEMS
Unemployment benefi ts are intended to provide
fi nancial security in case of unemployment and
to increase matching in the labour market by
giving jobseekers suffi cient time to fi nd suitable
jobs that match their abilities. However, by
raising workers’ reservation wages (i.e. the
wage level below which jobs are rejected), high
unemployment benefi t levels and long benefi t
durations tend to reduce the unemployed
person’s job search intensity and willingness to
accept job offers, and are thus likely to result in
an increased incidence and duration of
unemployment (for a survey of the empirical
literature, see OECD, 2006b, Chapter 3).72
As indicated in Table 12, unemployment benefi t
levels differ widely across euro area countries.
Over the last two decades, unemployment
benefi ts, as measured by the OECD summary
measure of unemployment benefi t entitlements,
have tended to increase in the euro area. In
more recent years, however, according to
this indicator, several countries have reduced
income support during unemployment,
in particular Germany, France and the
Netherlands. At the same time, in the majority
of euro area countries, the net replacement
rate 73 of unemployed persons with below-
average incomes was raised between 2001 and
2005, while for persons with average incomes,
Prepared by N. Leiner-Killinger.71
Furthermore, Genre et al. (2007) fi nd that female labour market 72
participation declines with the size of unemployment benefi t.
Bertola et al. (2002) show that unemployment benefi ts increase
prime-age employment relative to employment of young workers.
Centeno and Novo (2007) distinguish between a substitution and
an income effect generated by unemployment benefi ts. They fi nd
that an extension of the entitlement period in Portugal highlights
the importance of the income effect on labour supply, which
mitigates the distortionary nature of unemployment benefi ts. The
results indicate, however, that the most constrained individuals
benefi t the least from the extension of the entitlement period.
The ratio of an individual’s (or a given populations’ (average)) 73
net-income when unemployed in a given time period and the
(average) net pre-unemployment income when employed in a
given time period.
47ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
reductions and increases in the replacement
rate nearly balanced at the euro area level.
Over the past decade, the vast majority of
euro area countries tackled unemployment
benefi t administration by, for example,
tightening work availability conditions,
shortening the duration of benefi t payments
or tightening the eligibility conditions for
unemployment benefi t receipt (see Table 12).
In addition, several countries introduced in-
work benefi ts for workers working at low
wages (see OECD, 2006b, Chapter 3). On
average, these measures are found to have
supported employment creation by, inter
alia, raising the participation rate of persons
with low levels of education in particular
(see Chapter 3.1), and perhaps also by
supporting moderate wage increases, thereby
enhancing labour demand (see Box 5 for
reform experiences in Ireland and Germany).
4.1.2 LABOUR TAXES
Labour taxes, including income taxes and social
security contributions, affect labour supply,
equilibrium employment and hours worked, as
they drive a wedge between the marginal
productivity of labour and net income received.
In the presence of relatively high unemployment
insurance (i.e. net replacement rates), they may
have a particularly detrimental effect on
employment of workers with low incomes.74, 75
The effect of higher labour taxes (all else equal) on individual 74
labour supply is a priori ambiguous. On the one hand, higher
labour taxes may decrease disposable income, therefore
decreasing leisure and increasing the supply of labour (income
effect). On the other hand, labour supply may decrease, increasing
leisure, which is now cheaper (substitution effect). However, if
in general equilibrium, taxes are used to fund benefi ts such as
unemployment insurance. Negative income effects following
from higher labour taxes would tend to be cancelled out by the
positive income effect for benefi t recipients, while the substitution
effect goes in the same direction – reducing labour supply.
See e.g. Carone and Salomäki (2001).75
Table 12 Reforms of unemployment benefit systems in euro area countries
OECD summary measure of benefi t entitlements 1)
Net replacement rates 2) Reform policies 1994-2006 3)
67% of APW 100% of APW Shorter benefi t
duration
Tighter work availability conditions 4)
Tighter eligibility
conditions 5) level2005
change (p.p)1985-2005
change (p.p)2001-2005
level2005
change (p.p)2001-2005
level2005
change (p.p)2001-2005
Belgium 40.9 -2.2 2.4 71.0 2.0 56.0 1.0 + + +
Germany 24.2 -4.1 -5.2 78.0 -3.0 73.0 -2.0 + +
Ireland 33.7 5.4 4.0 70.0 3.0 59.0 5.0 +
Greece 13.3 6.1 0.2 65.0 2.0 47.0 3.0
Spain 36.0 1.6 -0.5 77.0 0.0 75.0 -1.0 - +
France 39.0 4.6 -4.5 83.0 0.0 67.0 -4.0 + +
Italy 32.5 32.1 -1.6 64.0 7.0 70.0 8.0 + +
Luxembourg 26.7 n.a. 0.0 90.0 1.0 89.0 0.0 +
Netherlands 35.3 -19.3 -17.5 86.0 1.0 70.0 0.0 + + +
Austria 31.9 2.5 0.3 72.0 -1.0 68.0 -1.0 + +
Portugal 39.5 17.9 -1.7 85.0 9.0 77.0 0.0 + + -
Slovenia n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Finland 35.3 0.9 0.4 85.0 -2.0 76.0 -4.0 + +
Euro area 32.3 4.1 -2.0 77.2 1.6 68.9 0.4
Denmark 48.9 -4.2 -2.0 91.0 0.0 75.0 -1.0 + + +
Sweden 23.8 -4.1 0.2 89.0 -1.0 69.0 -3.0 + +
United
Kingdom 15.6 -5.1 -1.0 70.0 18.0 60.0 16.0 +
United States 13.5 -1.2 0.0 49.0 -2.0 56.0 -3.0
Sources: OECD (2006b) Employment Outlook, Chapter 3, OECD (2007b) Benefi ts and wages.Notes: Unweighted averages for the euro area. n.a. not available.1) The OECD summary measure of benefi t entitlements is supposed to measure the overall generosity of unemployment benefi t systems. It is defi ned as the average of the gross unemployment benefi t replacement rates for two earnings levels, three family situations and three durations of unemployment.2) Net replacement rates are defi ned after tax in the initial phase of unemployment but following any waiting period for a one-earner couple with two children aged 4 and 6. The pre-unemployment income situation relates to 67% (100%) of the average production worker wage (for further details, see www.oecd.org/els/social/workincentives).3) Evaluations are based on OECD (2006b) and NCB assessments. + (-) indicates an increase (decrease) in the respective indicator. Assessments for Denmark, Sweden, the United Kingdom and the United States up to 2004 only.4) Including tighter requirements for being available for work when offered a job. 2005 for Austria.5) Including tighter eligibility requirements for certain groups of persons, most often increases in the minimum period of insured employment required to receive unemployment benefi ts.
48ECB
Occasional Paper No 87
June 2008
Moreover, the negative impact of social security
contributions on labour supply may be stronger,
the smaller the perceived link between workers’
social security contributions and benefi t
entitlements (‘perception effect’).76 Bassanini
and Duval (2006) and Silva (2005) document
that high tax wedges reduce employment. This
is supported by Nickell and Layard (1999).
Evidence also suggests that the impact of
taxation on labour supply may depend on the
way in which the public sector uses tax revenues
(see, e.g. Rogerson, 2007).
Table 13 surveys selected tax wedge
(the difference between employers’ cost of
employing workers and employees’ take-
home pay after tax) indicators for various
family situations. It shows that tax wedge
levels vary widely across euro area countries,
with differences amounting to more than
30 percentage points. In the majority of euro area
countries, tax wedges were reduced between
2001 and 2006, with the strongest reductions
taking place in Ireland (see also Box 5), which
has the lowest tax wedge levels across euro
area countries. The modest euro area-wide tax
wedge reductions between 2001 and 2006 tend
to underestimate long-term reform efforts, as
many countries undertook reforms lowering
labour taxes somewhat earlier.77 Overall,
there seems to be a tendency across euro area
See OECD (2007e), Chapter 4, pp. 170 for a discussion of the 76
literature.
For example, taking into account the year 2000 reduces the euro 77
area average tax wedges signifi cantly, i.e. by – 1.4 p.p for single
earners at 67% of the average worker wage, by – 1.2 p.p for a
one-earner married couple at 100% of the average worker wage
and by – 1.8 p.p for a two-earner married couple at 100% and
33% of the average worker wage.
Table 13 Changes in labour taxes in the euro area countries
Tax wedges 1) Unemployment trap 2) Low wage trap 2)
Single earner (67% of average worker wage)
One earner married couple (100%)
Two earner married couple (100%, 33%)
level 2006
change (p.p.) 2001-2006
level 2006
change (p.p.)2001-2006
level 2006
change (p.p.) 2001-2006
level 2005
change (p.p.)2001-2005
level 2006
change (p.p.)2001-2006
Belgium 49.1 -1.6 40.1 -2.5 41.0 -3.0 83.0 -3.0 58.0 2.0
Germany 47.4 -0.3 36.2 -0.6 41.5 -0.5 75.0 0.0 51.0 -2.0
Ireland 16.3 -1.1 2.3 -10.5 8.9 -7.9 76.0 3.0 53.0 7.0
Greece 35.4 0.3 41.5 1.8 40.0 1.5 59.0 3.0 19.0 1.0
Spain 35.9 0.6 33.6 0.9 35.4 0.2 80.0 0.0 26.0 2.0
France 44.5 -3.1 42.0 1.5 40.0 -0.6 81.0 0.0 35.0 -6.0
Italy 41.5 -1.2 35.1 -2.0 37.9 -2.5 71.0 12.0 33.0 4.0
Luxembourg 30.6 -0.6 13.0 -1.0 17.6 -0.7 88.0 0.0 51.0 8.0
Netherlands 40.6 1.7 37.0 8.8 36.8 5.6 86.0 7.0 70.0 5.0
Austria 43.5 0.6 36.9 2.0 37.7 1.9 67.0 0.0 37.0 2.0
Portugal 31.7 -0.5 26.6 -0.5 27.9 -0.3 81.0 0.0 20.0 -1.0
Slovenia n.a. n.a. n.a. n.a. n.a. n.a. 94.0 13.5 67.0 32.1
Finland 38.9 -2.5 38.0 -1.5 36.5 -1.8 76.0 -4.0 61.0 5.0
Euro area 38.0 -0.6 31.9 -0.3 33.4 -0.7 78.4 3.0 43.3 4.5
excl. SI 2.4 excl. SI 4.5
Denmark 39.3 -1.2 29.5 -1.1 34.4 -1.2 91.0 -1.0 82.0 -2.0
Sweden 46.0 -1.8 41.8 -1.1 41.7 -1.7 87.0 0.0 55.0 5.0
United
Kingdom 30.4 2.3 27.8 2.7 25.8 2.8 68.0 0.0 58.0 0.0
United States 26.4 -0.5 11.7 -3.1 19.3 -2.2 70.0 0.0 32.0 -2.0
Sources: OECD (2006d), “Taxing wages 2005-2006” and Eurostat Structural indicators database.Notes: Unweighted averages for the euro area. n.a. not available.1) The tax wedge captures income tax plus employee and employer social security contributions less cash benefi ts as a percentage of labour costs. It is displayed here for a single person without children with 67% of the average worker wage, for a one-earner married couple with two children aged 4 and 6 at 100% of the average worker wage and a two-earner married couple with two children, where one earner has 100% of the average worker wage and the other 33%. Ireland is based on the old OECD defi nition of earnings.2) The “unemployment trap” is defi ned as the percentage of gross earnings taxed away through higher taxes and social security contributions as well as benefi t withdrawal when an unemployed person takes up a job. It is measured here for a single person without children with 67% of the average earnings of a full-time production worker in the manufacturing industry.3) The “low wage trap” is defi ned as the percentage of gross earnings taxed away by higher taxes and reduced benefi ts when taking up a higher paid job. It is measured here for a single person without children, moving from 33% to 67% of the average earnings of a production worker.
49ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
countries to focus reductions in tax wedges on
persons with low earnings and on two-earner
married couples, which may have supported the
increase in employment of workers with low
earnings and of women observed in the recent
past (see Section 3.2.1).
However, as indicated by the high levels and
sometimes even unfavourable developments
in the unemployment and low-wage-trap
indicators across euro area countries displayed
in Table 13, the fi nancial rewards for moving
from unemployment to employment and from
lower to higher earnings remained rather low
between 2001 and 2005.
High levels of labour taxation may also provide
incentives to engage in non-employment
activities such as home production. Bertola et
al. (2002), for example, document that the
employment differential between females and
males is positively affected by labour taxes, as
higher taxes lead women to engage more in
home production.78 In addition, high levels of
taxes on income and value added (see Table 14),
and high social security contributions may lead
to more labour being supplied in the shadow
economy and thus often in activities with low
productivity. In a differentiated Value Added
Tax (VAT) rate system, a lower VAT may prove
particularly supportive of labour supply in
sectors whose services are easily substituted for
do-it-yourself or work in the underground
economy.79 In addition, according to the
literature, such incentives for involvement in the
countries’ shadow economies further arise from
the desire to circumvent high legal labour
standards, such as, inter alia, certain safety
standards or high minimum wages, as well as
product market regulations such as, e.g. time-
consuming administrative procedures.80
Tax and benefi t systems also affect incentives to
build up human capital, e.g. through subsidies
paid for participation in education, as discussed
in Section 4.4.
See also Burda et al. (2006) and Freeman and Schettkat (2005). 78
In turn, Genre et al. (2005) do not fi nd a signifi cant effect of
labour taxes on participation.
For a discussion, see Copenhagen Economics (2007). Some 79
euro area countries have, over the past two decades, raised
their standard Value Added Tax (VAT) rate, although standard
rates and the levels of reduced rates differ signifi cantly across
countries (see Table 14).
See Schneider (2006) for a survey. In a narrow sense, the 80
shadow economy may be defi ned as including market-based
production of goods and services that are deliberately concealed
from public authorities.
Table 14 Changes in Value Added Tax rates in the euro area countries
Standard VAT rateslevel change (p.p.) change (p.p.)2007 1984-2007 2000-2007
Belgium 21.0 2.0 2.0
Germany 19.0 5.0 3.0
Ireland 21.0 -2.0 0.0
Greece 19.0 n.a. 1.0
Spain 16.0 n.a. 0.0
France 19.6 1.0 -1.0
Italy 20.0 2.0 0.0
Luxembourg 15.0 3.0 0.0
Netherlands 19.0 0.0 1.5
Austria 20.0 0.0 0.0
Portugal 21.0 4.0 4.0
Finland 22.0 n.a. 0.0
Slovenia 20.0 1.0 n.a
Euro area 19.4 1.6 0.9
Denmark 25.0 3.0 0.0
Sweden 25.0 1.5 0.0
United Kingdom 17.5 2.5 0.0
Sources: OECD. Notes: Unweighted averages for the euro area. n.a. not applicable.
50ECB
Occasional Paper No 87
June 2008
Box 5
RECENT REFORMS OF TAX AND BENEFIT SYSTEMS. CASE STUDY: IRELAND AND GERMANY 1
Recent tax reforms implemented in Ireland have attempted to reduce the negative effects of high
marginal tax rates on labour supply. Reforms instigated in Germany between 2003 and 2005
instead sought to reduce unemployment benefi ts, tighten work availability requirements and
increase the attractiveness of employment. As these two examples show, reducing disincentives
to work, such as high marginal tax rates, high unemployment benefi ts and weak work availability
requirements can stimulate labour supply and employment.
The effect of taxation in Ireland
The tax wedge on labour was highest in Ireland in the late 1980s, fell back during the 1990s, and is
now among the lowest in the OECD countries (Nickell, 2003) and the euro area (see Section 4.1.2
above). Chief amongst these changes was the reduction in the marginal personal taxation
rate from 65% in 1983 to 41% in 2006. A narrower measure of the tax wedge that excludes
consumption taxes declined by 18.3% between 1986 and 2003 – the largest reduction recorded
in the OECD.
Higher taxes on labour contribute to higher unemployment and lower employment by reducing
labour demand (moving employers back up their demand curve) and labour supply (by moving
individuals back up their supply curves). The marked fall in these wedges during the 1990s can
be taken to have contributed to Ireland’s exceptional rate of employment creation. Research by
Callan, Van Soest and Walsh (2003) has explored the responsiveness of Irish labour supply to
various changes in the income tax code. In general, the results across different forms of tax cuts
were quite similar in many respects. However, the response of married women to a top rate tax
cut or to band-widening was more than twice as strong as that of men, and more than twice as
large as women’s response to a standard rate tax cut or allowance increase.
More recently, Callan, Van Soest and Walsh (2007) have examined the effects of increased
independence in the taxation of married couples. In 2000, the Irish tax system moved
towards greater independence in the tax treatment of couples in what has been termed an
“individualisation” of the standard rate tax band. This involved restricting the extent to which
tax bands are transferable between spouses. Callan et al. (2007) show that the impact of the
change in the tax treatment of couples on married women’s participation in the labour market
is substantially greater than cutting tax rates or increasing tax-free allowances. They estimate
that increased individualisation of tax increased married women’s participation rates by
2-3 percentage points. This increase is somewhat larger than that found by Steiner and Wrohlich
(2006) for Germany for a similar change in the tax treatment of couples.
Recent labour market reforms in Germany
Germany implemented major labour market reforms in 2003, 2004 and 2005, especially with
the so-called Hartz Acts I to IV (for an overview of these acts, see Deutsche Bundesbank, 2006,
pp. 79-83). The reform strategy included: improving employment services and redesigning active
labour market policy measures (Deutsche Bundesbank, 2006, p. 66); activating the unemployed
1 Prepared by K. Mc-Quinn and H. Stahl.
51ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
and reducing unemployment benefi t duration; tightening work availability requirements; and
stimulating labour demand by deregulating segments of the labour market (Eichhorst and
Zimmermann, 2007; and Jacobi and Kluve, 2007).
Prior to the reform, active labour market policy measures combined with a generous benefi t
system (unemployment benefi ts amounted to 67% of the last net labour income) created strong
disincentives towards work in Germany. After entitlement to unemployment benefi t expired (after
between 6 and 32 months, depending on previous employment duration and age), unemployment
assistance was paid for an unlimited amount of time and still amounted to 57% of previous net
labour income. For a single blue-collar worker with average income in western Germany this
would have been €833 per month in 2006 in current prices.
Hartz I and II redesigned the measures for recipients of unemployment benefi ts and unemployment
assistance and many of the existing active measures, promoted low-paid part-time employment
by waiving social security contributions and deregulated temporary employment. The ineffi cient
public employment service was completely reorganised under Hartz III. Hartz IV replaced
income-related unemployment assistance with the means-tested unemployment benefi t II in
2005.2 For a single person, unemployment benefi t II amounted to about €695 per month in 2006,
net of social contributions and taxes. The unemployment benefi t I duration entitlement was
shortened to 12 months for most of the unemployed and 18 months for the elderly (> 55 years)
at the beginning of 2006. Additionally, incentives for older persons to participate in the labour
force have been strengthened.
A thorough programme evaluation is mandatory under the Hartz Acts. Micro-evaluations of
Hartz I-III indicate that the reforms of the Federal Employment Agency had some positive
effects on employment, including the self-employed.3, 4 Training programmes in combination
with training vouchers for fi rms improved employment opportunities (Schneider et al., 2007).
However, publicly-sponsored job creation schemes and public-fi nanced work agencies for the
unemployed were found to have negative effects on employment. It is too early for a conclusive
assessment of the 2003 labour market reform act and Hartz IV. However, in a recent survey, fi rms
report an increased search intensity by the unemployed, along with reduced reservation wages
(Kettner and Rebien, 2007). Moreover, unemployment fell substantially in Germany after Hartz
IV was introduced (from 2005 onwards). The effect of even small reductions in replacement
rates on employment can be sizeable.
In terms of spending, in 2006, two-thirds of the measures have yet to be evaluated. One clear
outcome of the evaluations is that the number of measures of labour market policy, which amounted
to between 60 and 80, has to be drastically reduced. Marginal tax rates for unemployed persons
moving into employment are still very high. However, since every household whose income
falls below a certain threshold qualifi es for basic allowances, reducing the benefi t-qualifi cation
level would be diffi cult. Other measures, such as “workfare”, seem more promising.
2 Income-related unemployment assistance prior to the Hartz reforms was also means-tested, but conditions were weak and referred only
to partner income.
3 The promotion of self-employment through the provision of a benefi t to unemployed workers wishing to set up their own business (paid
for 6 months and equal to the unemployment benefi t that the recipient would otherwise have received - for further details see Jacobi
and Kluve, 2007) seems to have had long-lasting positive effects on employment. Caliendo et al. (2007) report that the probability of
not being unemployed after 28 months for men in Eastern Germany is 24% higher than for comparable men that have not received this
type of benefi t. For women in western Germany, this probability is 18% higher.
4 However the current German Government is in the process of reversing some of the reforms undertaken, inter alia by increasing the
duration of the higher unemployment benefi ts for older unemployed people
52ECB
Occasional Paper No 87
June 2008
4.1.3 PENSION SYSTEMS
Pension benefi t entitlements and their impacts
on labour supply differ widely across euro area
countries (e.g. OECD, 2007g; Blöndal and
Scarpetta, 1999; Martins, Novo and Portugal,
2007). Overall, studies point to signifi cant
negative relationships between pension benefi t
entitlement and early retirement access on the
labour supply of older workers (see, e.g. Duval,
2003 and 2006). As Table 15 indicates, the
portion of pre-retirement incomes that is
replaced by public pensions tends to be higher
for low-income workers.81 In 2004, in Greece
and Luxembourg, net pension entitlements for
low-income workers with 50% of average
earnings were even higher than previous income
from work, with net replacement rates exceeding
100%. Several reforms have been implemented
to reduce the fi nancial incentives for retiring
early. For example, for the euro area as a whole,
the implicit tax rate on continued work
embedded in early retirement schemes, which
becomes effective when a person decides to
extend employment from 60 to 65, declined by
18.2 percentage points between 1993 and 2003
but remains high.82 The reform strategies
enacted to increase working incentives for older
workers differ across euro area countries, with
most euro area countries increasing the statutory
retirement age, reducing the level of
unemployment benefi ts for older workers or
lowering the fi nancial incentives for retiring
earlier, e.g. pension levels and increases
(see Table 15 and also Box 6). Together, these
measures have been decisive in increasing the
labour supply of workers aged 55-64, as
indicated by the rise in this group’s labour force
participation rate over the last decade, as
identifi ed in Chapter 3 (see Table 2).
Further efforts to orientate tax and benefi t
systems towards increasing working incentives
and labour supply are needed. As regards
unemployment benefi t systems, these could
include improvements in benefi t administration
as well as adjustments in income support paid to
the unemployed where it reduces incentives to
search for work. Labour taxes need to be further
reduced. This can be achieved by restraining
government expenditures (e.g. through increased
effi ciency) and possibly by a partial shift from
social contributions and taxes on income to taxes
on consumption and other taxes, taking into
account redistributive effects. However, given
the impact of indirect taxes on shadow economy
workers, an overall tax reduction would most
likely be more effective for employment than a
simple shift in the tax burden for taxes on labour
to indirect taxes. As regards pension systems,
where necessary, early retirement incentives
need to be further reduced and the link between
social security contributions and benefi ts needs
to be strengthened.
Generally, a comprehensive approach to
increasing labour supply with mutually
reinforcing measures is decisive for increasing
labour supply and employment.83 By increasing
GDP, and thereby tax revenues and social
security contributions, such an approach allows
countries with sound structural fi scal positions
to further lower taxes on labour, thereby creating
a virtuous circle of raising labour supply.
Net replacement rates are estimated on the assumption that 81
individuals enter the labour market at age 20 and work until the
standard retirement age. See OECD (2007g), p.36.
Basically, the implicit tax rate on continued work measures the 82
amount of pension foregone (net of any additional contributions
paid) when an older worker decides to postpone retirement.
Since by retiring later, workers receive a pension for a shorter
period, to be incentive neutral this would need to be compensated
by a higher yearly pension in the future. If the pension increase
for retiring later falls short of compensating for one lost year
of pension, the system is said to favour early retirement. The
implicit tax rate is the amount of pension “lost” due to later
retirement (in present value terms), relative to labour income.
The implementation of tax and benefi t policies across euro area 83
countries often takes the form of policy packages, combining
either various reforms of tax and benefi t systems or linking such
reforms to reforms of other labour market institutions. The latter
is the case, for example, within the currently discussed fl exicurity
approach. Broadly speaking, this approach is supposed to combine
elements of labour market fl exibility with elements of income
security. A ‘suffi cient’ fl exibility in employees’ contractual
arrangements is meant to allow fi rms and employees to cope with
(structural) change, while employees shall be eligible to receive
an ‘adequate’ income in periods of joblessness that may arise
from the increased labour market fl exibility. Overall, the concept
of fl exicurity is intended to take account of trade-offs between
social transfers, employment protection legislation, active labour
market policies, as well as life-long learning strategies within an
integrated approach. It is decisive that such reform packages exert
a positive impact on labour supply by increasing overall labour
market fl exibility without giving rise to budgetary distortions
through higher taxes.
53ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
Table 15 Reforms of pension systems in euro area countries
Net replacement rates 1) Implicit tax rates 2) Policy reforms 1994-2006 3)
50% of av. income level
100% of av. income level level change (p.p)
Reduced early
retirement 4)
Lower unemployment
benefi ts 5)
Financial incentives for
later retirement 6)2004 2004 2003 1993-2003
Belgium 77.3 63.0 76.5 5.9 + + +
Germany 53.4 58.0 39.0 -14.2 + + +
Ireland 65.8 38.5 37.3 -0.6
Greece 113.6 110.1 n.a. n.a. -
Spain 82.0 84.5 33.6 22.1 + [+;-]
France 78.4 63.1 50.6 -32.0 + -' + +
Italy 81.8 77.9 20.6 -79.8 + + +
Luxembourg 107.6 96.2 75.3 -3.5 + +
Netherlands 97.0 96.8 29.5 -59.8 + + +
Austria 90.4 90.9 65.6 n.a. + +
Portugal 81.6 69.2 76.5 0.9 + + +
Slovenia n.a. n.a. n.a. n.a.
Finland 77.4 68.8 59.4 -21.4 + + +
Euro area 83.9 76.4 49.8 -18.2
Denmark 132.7 86.7 n.a. n.a. + + +
Sweden 81.4 64.0 35.7 14.3 + [+;-]
United Kingdom 66.1 41.1 26.4 6.2
United States 67.4 52.4 12.8 6.5 [+;-]
Sources: OECD (2006b) Employment Outlook, Chapter 3 and OECD (2007g) “Pensions at a glance”; N. Brandt, J.-M. Bruniaux andR. Duval (2005), “Assessing the OECD Jobs Strategy: Past Development and Reforms” OECD Economics Department Working Paper No 429.Notes: Unweighted averages for the euro area. n.a. not available. Changes in the replacement rate over time are not available.1) The net replacement rate is defi ned as the individual net pension entitlement divided by net pre-retirement earnings, taking account of personal income taxes and social security contributions paid by workers and pensioners. It is displayed at 50% and 100% of average earnings, respectively. Displayed for retirement in 2004.2) The implicit tax on continued work is defi ned as the average annual change in pension/social wealth (i.e. the present value of the future stream of pensions/social benefi ts), net of additional contributions paid, resulting from a decision to postpone retirement from age 60 to age 65. The calculations are made for a single worker with average earnings. For 2003, they refl ect the steady-state of currently legislated systems thus taking account of recent reforms and their impact on future pension streams, which in some cases will take several decades. The fi gure for Belgium refers to the period 1995-2003.3) Evaluations are based on OECD (2006b, 2007g) and NCB assessments. + (-) indicates an increase (decrease) in the respective indicator or simultaneous measures acting in both directions. Assessments for Denmark, Sweden, the United Kingdom and the United States up to 2004 only.4) Includes tightened eligibility to early retirement (e.g. through increase in retirement age for early retirement).5) Includes unemployment benefi ts for older persons seeking a job with an extended duration of unemployment benefi t receipt.6) Includes actuarial adjustments to early or late receipt of pensions or fi nancial incentives to retire later (e.g. bonuses)
Box 6
ENHANCING THE PARTICIPATION RATE OF OLDER WORKERS: THE NEED FOR A COMPREHENSIVE
STRATEGY 1
Despite the fact that in most euro area countries the statutory retirement age is currently 65
for both men and women (see Table and Social Security Administration, 2006),2 in 2006, the
participation rate of individuals aged 55 to 64 stood, as for younger individuals, at around 44%.
This low rate of participation for older workers refl ects cohort effects (for women), but is also
due to economic and institutional factors. In fact, the participation rate of men between 55 and
1 Prepared by D. Nicolitsas.
2 See Table for details on the statutory pensionable age in 2005. The only euro area countries in which the statutory pensionable age
for men was not 65 in 2005 were France (60) and Slovenia (63). The euro area countries in which the statutory pensionable age for
women was less than 65 in 2005 were Belgium (64), Greece (60 but only for those women who joined the labour force prior to 1993),
France (60), Italy (60), Austria (60) and Slovenia (61).
54ECB
Occasional Paper No 87
June 2008
64 years of age in the euro area is lower today than two and a half decades ago, while the average
effective retirement age for both men and women is lower today than in the 1980s.3
The signifi cant discrepancies between countries in the effective retirement age (see Table)
refl ect, inter alia, cross-country differences in the level and composition of economic activity,4
and in the extent of self-employment. The main factors, however, behind these discrepancies are
differences in: the statutory retirement age, early retirement schemes, restrictions on employment
during retirement, and replacement rates.5
The decline in the average effective retirement age is mainly due to the fact that in the 1980s a
number of countries tried to tackle unemployment, especially that of the young, by introducing
early retirement schemes (e.g. Belgium, Germany, Austria) and made disability benefi ts more
generous (e.g. Netherlands). At the same time, the increase in the unemployment rate amongst
older people led them to withdraw from the labour market (known as the discouraged worker
effect).
The main reason for which individuals take advantage of early retirement schemes are the high
implicit tax rates on continued work. As can be seen from Table 13, implicit tax rates are positive
in all countries and very high in some. In other words, Blöndal and Scarpetta (1999) estimate
that in the 1990s in a number of OECD countries a 55 year-old man could expect no increase in
his pension if he continued working for another 10 years – a situation in sharp contrast with what
prevailed in the 1960s.
More recently, given other developments (e.g. population ageing and the need for fi scal
consolidation), it became clear that while measures such as early retirement schemes and more
generous unemployment and disability benefi ts were not effective in lowering unemployment,6
they could, ceteris paribus, lead to a reduction in output and mount pressure on public fi nances by
reducing the size of the aggregate labour force. For this reason, a number of countries reversed
such measures and started to give employers incentives to increase the number of older workers,
and workers incentives to stay at work longer.7
Limiting the use of early retirement schemes will have wider benefi ts than just helping public
fi nances. More specifi cally, such measures could contribute to increased training activities by
lengthening the pay-back period for investment in human capital or by encouraging older workers
to weather a temporary demand shock without withdrawing permanently from the labour market
(see, inter alia, Duval, 2003). Moreover, staying in work can help older people remain integrated
in society. Despite the benefi cial effects expected as a result of the measures that have already
been planned, and the fact that the participation rate of older workers is expected to increase due
3 See Blöndal and Scarpetta (1999) for evidence until the mid-1990s and Romans (2007) for evidence on the more recent period.
4 Individuals from certain sectors in which work is more arduous (e.g. construction, manufacturing, mining and quarrying and
transportation) generally have earlier retirement dates.
5 Differences in the decision to retire early also differ across individuals depending on, inter alia, their education level; in general, there
is a negative correlation between early retirement and education level.
6 See OECD (2006d) for an illustration of the point that in 2004, countries in which participation rates of older workers were low also
had high unemployment rates (Chart 9, p.132).
7 For example, in Germany the statutory retirement age will, between 2012 and 2029, progressively increase from 65 to 67 starting
with the age cohort born in 1947; in Austria early retirement schemes are to be abolished with effect from 2017; in France a pension
supplement is given to people over the age of 60 who, despite the fact that they could earn a full pension, decide to stay on at work;
social security contributions for older employees have decreased in Belgium; in Portugal, the pension scheme reform enacted in 2007
included a bigger fi nancial penalty for early retirement – an increase from 4.5% to 6% for every year prior to the legal retirement age –
and introduced a “sustainability factor” that will make the calculation of new pensions conditional on life expectancy at 65.
55ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
to the improvement in the educational level of the population as a whole and due to the increased
importance of the service sector, there is still need for further action. This is prompted by the
rapid expected deterioration in the old-age dependency ratio 8 (from around 25% in the euro area
in 2006 to around 35% in 2025 and 55% in 2050).9 In addition, there is still a discrepancy of
around 8 percentage points between the euro area employment rate of individuals aged 50-64
and the target of 50% for 2010 set by the 2001 Stockholm European Council. The signifi cant
cross-country differences and past experience with policy measures suggest that for the adopted
measures to succeed, it will take a comprehensive strategy that incorporates many of the special
features needed to promote the activation of older workers, not least of which is the need for
life-long learning (see, inter alia, Employment Committee, 2007). Such comprehensive reform
should also take into account other measures which may currently dampen incentives for further
training and render older employees less employable, for example those arising from the rigidity
of age-earnings profi les.
8 The old-age dependency ratio is defi ned as the population over 64 years old as a percentage of the working age population.
9 Figures from Maddaloni et al. (2006).
Statutory pensionable age and average exit age from work by gender (2005)
Statutory pensionable age
(1)
Early pensionable age
(2)
Average exit age
(3)
Statutory pensionable age
(4)
Early pensionable age
(5)
Average exit age
(6)Men Women
Belgium 65 60 61.6 64 (65) 1) 60 59.6
Germany 65 (67) 2) 63 61.4 3) 65 (67) 2) 63 61.13 3)
Ireland 65 - 63.6 65 - 64.6
Greece 4) 65 60 62.5 60 4 55 61
Spain 65 - 62 65 - 62.8
France 60 - 58.5 60 - 59.1
Italy 65 - 60.7 60 - 58.8
Luxembourg 65 57 59.4 65 57 59.4
Netherlands 65 - 61.6 65 - 61.4
Austria 65 62.5 5) 60.3 60 5) 58.5 5) 59.4
Portugal 65 55 62.4 65 55 63.8
Slovenia 63 - n.a. 61 - n.a.
Finland 65 62 61.8 65 62 61.7
Euro area 6) 64.2 60.3 61.5 62.6 58.9 61.2
Denmark 65 60 61.2 65 60 60.7
Sweden 67 61 64.3 67 61 63
United Kingdom 65 (68) 7) 50 (55) 7) 63.4 60 (68) 7) 50 (55) 7) 61.9
United States 65 62 n.a. 65 62 n.a.
Sources: DG Employment, Active ageing country fact sheets (for columns: 1, 3, 4 and 6) and Social Security Administration, 2006 for columns 2 and 5. 1) The statutory retirement age will, from 2009, increase to 65.0 for women. 2) The statutory retirement age will gradually (between 2012 and 2029) increase to 67, starting with the age cohort born in 1947. 3) Data refers to 2004. 4) In Greece, data refer to the modal case of the main pension provider for private sector employees. Statutory retirement ages are considerably lower for public sector employees, while they are less generous for farmers and the self-employed. The statutory pensionable age is 65 years for women who joined the labour force since 1993. The median exit age for women was 58.4 in 2006. 5) Early retirement is being increased gradually (by one month every four months) until it equals the statutory age – i.e. until 2017. 6) Average for the countries for which data are presented in the Table. 7) The statutory pensionable age will increase from 65 to 68 by one year in each of the years 2024, 2034 and 2044; statutory pensionable age will be equalized for men and women in the period 2010-2020; and early retirement age for a non-state pension will increase from 50 to 55 by 2010.
56ECB
Occasional Paper No 87
June 2008
4.2 PARENTHOOD AND PARTICIPATION 84
As discussed in Chapter 3, the labour market
participation and employment rates of females
are still low compared with those of males.
According to recent calculations, euro area
GDP would be substantially increased if
female employment rates could be raised to
those of men (see Daly, 2007). Female labour
supply, that of parents in general, and fertility
developments (and thus long-term labour
supply and potential output) are substantially
infl uenced by so called “work/family
reconciliation” policies, namely childcare
opportunities, parental leave and part-time
work opportunities. These policies differ across
euro area countries and may consequently help
to explain differences in female labour supply
and fertility across countries. In general, it is
important to create incentive-compatible policy
frameworks that support both female labour
market participation and high fertility.
This section briefl y describes how these policies
affect female and parents’ labour supply and
suggests optimal policy designs for childcare,
parental leave and part-time work opportunities
for reconciling fertility and female labour
market participation.
4.2.1 FERTILITY AND FEMALE PARTICIPATION:
WHAT IS THE RELATIONSHIP?
Box 1 in Chapter 3 highlighted the importance
of fertility rates for future labour supply. Most
countries with high participation rates also have
relatively high fertility rates (such as Finland,
France and the Netherlands – see Chart 8). At
the same time, a number of countries are plagued
with both mediocre or low participation and low
fertility (Italy, Spain and Greece), while others
such as Germany and Austria have relatively
high participation rates, but low fertility rates.85
When comparing the situation in 1992 with that
of 2006, most countries retain a similar relative
position, despite the general increase in female
participation rates described in section 3.2.1.
One exception is the Netherlands (and to some
extent Ireland), which managed to move from
mediocre to become a top performer in both
categories.
Prepared by M. Ravanel and H. Schauman.84
See also Genre et al. (2005) and OECD (2001).85
Chart 8 Female participation, employment and fertility in euro area countries, 1992 and 2006 1
x-axis: participation and employment (in percentage); y-axis: fertility rate (average number of children per woman)
BE
DE
IE
GRES
FR
LUNL
PT
FIUK
IE
FIUKFR
BELU
NL PT
GR
DEES
1.0
1.2
1.4
1.6
1.8
2.0
2.2
1.0
1.2
1.4
1.6
1.8
2.0
2.2
80
participation rate (in percent)
employment rate (in percent)
denotes the average participation rate for the euro area
30 40 50 60 70
1992
1.0
1.2
1.4
1.6
1.8
2.0
2.2
1.0
1.2
1.4
1.6
1.8
2.0
2.2
7540 45 50 55 60 65 70
IEIE
BE
DEIT
GR
ES
LU NL
PTATAT
FI
IT
FI
SISI
UKUK
FR FR
BE
LUNL
PTGR
DEES
2006
denotes the average employment rate for the euro area
denotes the average fertility rate for the euro area
Source: Eurostat.Note:1) 2005 for Spain and Italy fertility fi gures.2) Countries for which no data were available for 1992 or before have been omitted.
57ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
These different country groups refl ect differences
in policies affecting labour market participation
and fertility across countries. Policies that aim to
simultaneously increase both fertility and labour
market participation rates must focus on two
issues: fi rst, increasing the labour supply of parents
that do not yet participate in the labour market; and
second, working females (or future parents) who
are considering the fertility decision. Here, the
opportunity cost of having a child, in terms of
labour market withdrawal and possible negative
consequences for the female career, may have a
particularly strong impact on the fertility rate of
highly skilled women. Different, non-exclusive,
systems can be implemented, including the
following three possible solutions – increasing the
availability and lowering the cost of child care,
parental leave and part-time work opportunities.
Interestingly, two of the success stories, in terms
of female employment rates, namely the
Netherlands and Finland, use very different policy
mixes. In the Netherlands, fl exible working hours
and a high percentage of part-time work in total
employment appear important for both high
participation and fertility (most children do not
attend full time childcare), while in Finland, high
full-time employment and fertility is supported by
a developed childcare and extensive parental leave
system (see Box 7). The next sections will consider
these systems and their effect on participation and
fertility decisions in greater detail.86
4.2.2 AVAILABILITY AND COST OF CHILDCARE
It is widely documented that well-designed
childcare opportunities have a positive impact on
female participation (see, for example Bassanini
and Duval, 2006) and allow parents to continue
with full- or part-time work after having a child.
However, the availability and fi nancial cost of
childcare systems are important. When childcare
provided through the market is too scarce or
too expensive, one parent, usually the mother,
often stays at home to take care of the children
until they go to school. This not only keeps the
parent away from market work during the leave,
but can also affect the parent’s employment
situation after the leave due to lack of relevant
professional experience. Table 16 provides
fi gures on the type of childcare system most
frequently used by working mothers in some
European countries. Eurostat (2007) fi nds that
when no market-based childcare system is used,
the main reason given by families across the
European Union is that it is too expensive. This
can lead low-paid employees to leave the labour
market, or to accept that a large proportion of
their wage will be re-channelled into childcare.87
Note that in Finland, the price paid for childcare
is linked to family income (see Box 7).
The European 86 ad hoc module of the 2005 LFS focused on
conciliation between family and professional life. A report
issued by Eurostat in 2007 called Reconciliation between work and family life gives very interesting insights on the availability
of those different solutions across the European Union. We will
refer to their main results, especially for childcare availability.
Increasing childcare opportunities is in line with one of 87
Barcelona targets. The European Council stated that “Member
States should remove disincentives to female labour force
participation and strive, taking into account the demand for
childcare facilities and in line with national patterns of provision,
to provide childcare by 2010 to at least 90% of children between
3 years old and the mandatory school age and at least 33% of
children under 3 years of age”.
Table 16 Principal type of childcare used by employed mothers during hours of work
(this refers to children under 15, where at least one child is under 6 years old, in percent)
Market child Family, friend,care system Mother Partner Neightbour
Belgium 66 3 6 25
Germany 51 6 23 20
Spain 37 22 11 30
France 46 13 11 30
Italy 35 8 20 37
United Kingdom 34 20 17 29
Total % in all 6 countries 40 15 15 30
Sources: Micheaux, S and O. Monso (2007), European LFS 2005 and ad hoc module 2005. Data are only available for the six countries presented. Note: “Market” child care system includes both public and private provision. The column "mother" includes the case where no alternative child care is available.
58ECB
Occasional Paper No 87
June 2008
It is important to recall that there are a number
of possible means available to lower the fees
charged for market childcare systems. These
include: increasing publicly provided and
fi nanced child care opportunities; subsidising
private provision directly; or introducing a
voucher system. Given the high costs of
childcare systems together with the need for a
balanced government budget, the effi cient use
of resources is crucial. The voucher system
supports a market-based childcare system with
the benefi ts of free competition and choice, but
has generally not been used in the euro area
countries so far. Private efforts and investments
associated with raising and educating children
have positive “external” effects for society (e.g.
in the form of future tax payments). These can
justify certain subsidies for families raising
children. However, direct child-related transfer
payments to parents may decrease incentives to
participate in the labour market. The Austrian
extension of the “Kindergeld” (child allowance)
scheme in 2002 is an example of such an effect 88
and may help to explain a negative relationship
between fertility and labour market participation
at the individual level.
For more information, see OECD (2007a). In 2002, the Austrian 88
government extended the period of entitlement for the so-called
“Kindergeld” scheme from 18 to 30 months. The maximum
duration can be extended by six months if the partner also
decides to take part in the program. Men, however, rarely
take the opportunity. The generosity of the scheme (it pays
approximately €430 per month) and the rather low threshold
for any work-related income induce most women to stay at
home for quite a long time. Empirical research has shown
that the Kindergeld scheme signifi cantly decreases women’s
participation rates. It has also been criticised on the grounds that
it is likely to lower women’s re-entry prospects in the labour
market. Critics also argue that the subsidy should be in the form
of childcare vouchers rather than a money transfer. Recently,
the Austrian government has enacted a reform to make the
Kindergeld more fl exible. Starting from 2008, parents may
chose between different lengths of payment and thus trade off
benefi t duration and monthly transfer payments.
Table 17 Length of parental leave and level of benefits
Country Maternity leave duration, weeks
Parental leave duration, weeks
% of salary during maternity leave
Paternal leave duration, days
Austria 16 100
Belgium 15 82 1)
Finland 17.5 26.5 70 2) 18
France 16 100 11
Germany 3) 14 60 (52) 100
Greece 17 100 7) 2
Ireland 18 80
Italy 21 80
Luxembourg 16 26 8) 100 2
The Netherlands 16 100
Portugal 4) 20 5 100
Slovenia 5) 15 37 100 15
Spain 6) 16 (10) 100 2
Sources: Social security programmes throughout the world, 2006 and information from the national central banks. The periods of leave listed here refer to leave granted in connection with childbirth, which usually includes a period of leave prior to and after the birth of a child. Maternity leave refers to the weeks of leave from work which can only be taken by mothers. Parental leave refers to leave that can be taken by either the mother or the father of a child. Paternal leave is optional leave that can be taken by the father in order to allow both parents to be on leave at the same time - the mother on maternity or parental leave and the father on paternal leave. In general, leave systems are complex. This table attempts to summarise the key points of the systems in place in each country. 1) For 30 days, 82% of salary; thereafter, 75%. 2) Payment up to a ceiling of €28,402 plus 40% of daily earnings for annual earnings between €28,404 and €43,698, and 25% of daily earnings for annual earnings of €43,699 or more. The minimum daily benefi t is €15.20. 3) At the beginning of 2007, the so called “Elterngeld” – a subsidy to parents who leave their job to take care of a child - was introduced. The maximum duration of 14 months requires that the leave period is shared between parents, with a minimum leave period of at least 2 months for each parent. Part-time work up to 30 hours a week is possible. 4) If the maximum leave period of 25 weeks of maternity and parental leave is opted for, the percentage of salary paid for the whole period falls to 80%. 5) The parental leave in Slovenia is known as a childcare benefi t. 6) Of the 16 weeks of parental leave, the mother must use 6; the remaining 10 can be used by either the mother or father. In 2007, paternal leave increased to 15 days. 7) 50% of salary is covered by the maternity benefi t, but mothers receive 100% of their salary while on maternity leave. 8) 13 weeks can be taken by the mother and 13 weeks by the father of a child up until the child’s 8th birthday. This is generally unpaid leave, although in certain collective agreement arrangements may be partly paid.
59ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
4.2.3 MATERNAL AND PARENTAL LEAVE
OPPORTUNITIES
Parental leave exists in every country of the
euro area; however, benefi ts and their
duration differ from one country to another
(see Table 17).89 Maternal and parental leave
schemes generally tend to positively affect the
labour supply of females and parents
(Jaumotte, 2003). Having the right to suspend a
work activity, with the option to come back to
work afterwards without suffering a wage loss
during the leave period, can have a positive
infl uence on the decision to have a child.
Furthermore, parental leave may help
redistribute the time needed for childcare more
evenly between parents. However, Genre et al
(2005) show that parental leave has a positive
effect on employment as long as the leave period
is for less than 10 months. In addition, parental
leave schemes have been found to incur a cost
in terms of (mainly female) career progression 90
and to the extent that they are fi nanced by fi rms,
may impose a cost on employers.91
4.2.4 PART-TIME WORK OPPORTUNITIES
Part-time work is a further policy option to
conciliate family and professional lives, under
certain conditions. Indeed, some evidence in the
literature points to a positive impact of part-time
work on employment 92 and Section 3.3.2 has
highlighted the increase in part-time work since
the early 1990s, which coincided with a strong
increase in female labour supply.
At fi rst sight, working part-time seems to offer
a popular opportunity to arrange family and
work, since women choose to work part-time
far more frequently (as can be seen in Chart 9).
However, although part-time work opportunities
increase labour market participation, this does
not necessarily result in higher employment
when measured in average weekly (or annual)
hours of work (see Section 3.3.2). As shown
in Chart 9, rates of involuntary part-time work
are relatively high for females.93 For parents,
high costs or unavailability of childcare may
force workers to work part time rather than
full time. Furthermore, part-time work is
estimated to have a negative impact on career
progression.94 Balancing the rights of part-time
and full-time workers, allowing fl exibility for
both fi rms and employees through the equal
access of all working groups to part-time work
and discouraging the negative stigma that can
sometimes be associated with working part
time in some cases should be benefi cial for all
parties. Furthermore, favouring fl exible working
arrangements and teleworking could provide an
alternative to part-time work for reconciling
family and professional life.95
The information presented in Table 17 refers to parental leave 89
related to the birth of a new child, not leave for childcare more
generally.
See Mincer & Polachek (1974), Gronau & Weiss (1981) and 90
Stoiber (1990).
Possibly introducing a bias by fi rms against the hiring of 91
women.
Genre et al. (2005) document that part-time work positively 92
and signifi cantly affects the labour supply of the young, and
Pissarides et al. (2003), inter alia, emphasise the positive effects
of part-time work on female labour supply.
Generally, part-time workers have been shown to suffer 93
detrimental effects of their shorter working week on career
progression (see for instance International Labour Review
(1997), or Tam (1997)).
See for instance International Labour Review (1997) and Tam 94
(1997).
For example, in the Netherlands, fl exible working arrangements 95
and a low rate of (involuntary) part-time work coincide.
Chart 9 Part-time workers in euro area countries in 2006 as a percentage of total employment, including involuntary part-time.
(in percent)
0
10
20
30
40
50
60
70
80
0
10
20
30
40
50
60
70
80
part-time
including involuntary
mf mf mf mf mf mf mf mf mf mf mf mf mf mf mf
m males
f females
1 Euro area
2 Belgium
3 Germany
4 Ireland
5 Greece
6 Spain
7 France
8 Italy
9 Luxembourg
10 Netherlands
11 Austria
12 Portugal
13 Slovenia
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
14 Finland
15 United Kingdom
Source: Eurostat. Notes: Data on involuntary part-time work are not available for males in LU and the euro area. In these cases, only data on voluntary part-time employment as a percentage of total employment is shown (by the dotted histograms).
60ECB
Occasional Paper No 87
June 2008
Box 7
REFORM OF CHILDCARE AND FLEXIBLE WORKING ARRANGEMENTS. CASE STUDY: FRANCE AND FINLAND.1
This box examines one important determinant of female labour market participation, namely
the childcare and fl exible working arrangement schemes in two euro area countries, Finland and
France. Finland experiences the highest female participation rate in the euro area and a high
fertility rate. France also has very high fertility rates but has female participation rates below
65%. This box gives an overview of the Finnish system and presents the most recent reform, the
2004 PAJE-reform, in France.
In Finland, a record 67.3% of women worked outside the home in 2006, compared with 71.4%
of men. Most mothers of young children work full time and only 19.2% of working women work
part-time, compared with the euro area average of around 35%. Some of the explanations for
the high female employment rate can be found in a relatively successful mix of fl exible working
arrangements for the parents of young children and effi cient childcare services.
The childcare system in Finland started in 1973 with a law on childcare encouraging
municipalities to provide childcare for all children up to the age of 6. In contrast, from 1990
onwards, municipalities were obliged to arrange childcare for all children up to the age of 3, and
since 1996 the same is true for all children below school age (the year the child turns 7). Roughly
50% of children below school age use the municipality childcare service. Of these, 77% are in
full-time day care. Only 3.5% of all children in childcare attend private childcare. There are also
part-time childcare services and around-the-clock childcare for children whose parents work in
shifts. The cost of childcare borne by parents is linked to family income - the highest fee is €200
per month, falling with each additional child. Low-income families do not pay childcare fees.
The Finnish family leave system builds on the principle that both parents should have equal
opportunities to take part in the caring of a child. For each newborn child, family leave is a
maximum of 263 days. This is split into maternal leave (105 days) and parental leave (158 days).
In addition, there is a short paternal leave (18 days). The fi nancial support paid to parents during
the leave period is tied to regular income and is at least €15.20 per day. Paternal leave allows
the father to stay at home, together with the mother, when the child is born. The popularity of
paternal leave has grown steadily, with 69% of all fathers using their right to paternal leave in
2005. After the parental leave period, one parent is allowed to get an unpaid leave from his or
her job for nursing his or her child until the child turns 3. The monetary support for one child is
roughly 300 euros per month but grows with the number of children.
In France, the fertility rate is high (at 1.93 children per woman), but female labour market
participation remains signifi cantly below that of males. Before 2004, allowances aimed to
increase fertility by lowering the “cost” of having a child. In 2004, the allowances designed to
raise fertility were reformed in order to also raise female participation.
The French allowance scheme, PAJE, consists of a birth grant, a monthly allowance paid until
a child’s fi rst birthday and two allowances for the “free choice” to reduce working time (paid to
parents who stop or reduce their paid employment to take care of their child) 2 and of third-party
1 Prepared by M. Ravanel and H. Schauman.
2 A very similar system was implemented in Germany in 2007, where parents can, under certain conditions, receive an allowance.
61ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
4.3 IMMIGRATION POLICIES 96
As argued in Chapter 3, immigration can help to
fi ll skill shortages, contributing to the reduction
in wage pressures stemming from an increased
demand for particular skills. This suggests that
an effective immigration policy should take
into account the skills demanded in the labour
market. As discussed in Box 2, migrants can
bring skills which natives lack and cannot
acquire to a suffi ciently large scale in the short
run. Accordingly, a number of countries utilise
migration policies that select the type and amount
of immigrants let into a country over a specifi ed
time period, and design mechanisms that attempt
to facilitate the rapid integration of immigrants
with a more permanent status. Moreover, at least
in the short run, the entry of a sizeable number of
foreign workers may help to alleviate the reduction
in the size of the labour force in the near future
(see Box 1), provided that markets are suffi ciently
fl exible and policies suffi ciently effi cient to
ensure their smooth integration into the labour
market. However, there is general agreement that
immigration alone clearly cannot solve the longer-
term problems of Europe’s ageing population on
the pension and health care schemes.97
The contribution of immigration to labour
supply is regulated through immigration policies.
Traditionally, most EU governments have
selected the type of immigrants and length of
residence that suit their country’s needs. Usually
there is an annual limit (quota) for non-EU and
some EU immigrants who fulfi l a selection
criterion. As is evident from Chart 7, labour
demand has encouraged an increase in the quotas
since the beginning of the 1980s in euro area
countries.98 Furthermore, immigrants from the
EU15 99 have the right to freely move across the
euro area 100 and the EU enlargement in 2004
contributed to the increase in immigration within
the EU. Countries have tended to recognise rights
regarding asylum seekers, the marriage with a
Prepared by A. Lacuesta.96
European Commission (2002a). This conclusion is also evident 97
in the Council Opinions on the Stability Program Updates of
many European countries.
As it is further explained in OECD (2007e) (also see Box 2), 98
during the 1980s migrants generally concentrated in traditional
immigration countries (mainly Germany and Austria), while in
the 1990s, the number of immigrants increased the most in new
immigration countries (Spain, Ireland and Italy).
That is, Belgium, Germany, Ireland, Greece, Spain, France, 99
Italy, Luxembourg, the Netherlands, Austria, Portugal, Finland,
Denmark, Sweden and the United Kingdom.
The free movement of persons is one of the fundamental rights 100
guaranteed to EU citizens, and includes the right to work and
live in another EU Member State. Temporary restrictions on
the mobility of migrants to and from a new EU Member State
are possible for up to seven years following EU enlargement.
Such restrictions have been in place for some non-EU-15 EU
countries since the 2004 and 2007 EU enlargements. Moreover,
regulations such as minimum wages effectively undermine free
movement.
childcare (which pays part of a childminder’s wage). Since 2004, the aim has been to develop
participation through part-time work. Thus, the two allowances paid to part-time working parents
have been increased. This has had a positive effect on participation, but 60% of parents still choose
not to work at all. 98% of them are women, and 84% of these are blue collar workers. 37% of parents
who chose to stay at home say that another choice would have been too expensive (CNAF, 2006).
Thus, participation of low-paid mothers has not dramatically risen since the PAJE reform. Crèches
(in other countries also known as nurseries) remain the cheapest system. However, availability in
crèches is low and the opening hours do not cover a complete working day. Because availability is
so scarce, crèches are the principal childcare system for only 8% of children under 3 years of age.
In sum, the availability of affordable (e.g. subsidised) formal child care services and municipal
day care services, including crèches, in general has a positive impact on female participation
(Jaumotte, 2003; Gustafsson and Stafford, 1992; and Viitanen, 2005). Allowance packs such as
PAJE can, on the other hand, have negative effects on participation, especially of low-skilled
women. Increased availability of municipal day care services, or possibly a voucher system to
pay for private day care services, would allow for better work/family reconciliation and support
future female labour market participation.
62ECB
Occasional Paper No 87
June 2008
foreign partner, adoption of foreign children and
reunifi cation of a foreigner’s family. This last
mechanism is an endogenous source of growth for
migration. In traditional European immigration
countries, immigrants from family reunifi cation
account for more than half of the total annual
infl ow of immigrants (OECD, 2006b).101 In
addition, income differences between origin
and destination countries combined with the
fall in transportation costs have encouraged
illegal immigration. Increased cooperation
among Member states and with countries of
origin is needed to decrease illegal migration.
Furthermore, fi rms need to be discouraged from
illegal hiring of migrants (since a large fraction
of undocumented migrants are immigrants that
have over-stayed their legal residence rather
than new entrants; OECD, 2007e). In sum, the
abovementioned factors emphasise that active
immigration policies have some limitations
in immigrant selection; however, many euro
area countries have selected their immigrants
according to employment-based migration
policies described below.
Employment-based migration policies aim to
attract particular skills in short supply on the
national labour market. Usually employers are
responsible for hiring directly, strengthening the
link between immigration and the needs of the
national labour market through ensuring a job for
the immigrant upon arrival. Governments typically
have to approve a migrant’s entry following
consideration of the country’s labour market
situation.102 Since the system is intended to solve a
limited shortage, the visa awarded is usually
temporary, although in some cases it is possible to
apply for permanent residency after several years
of residence in the destination country. The system
works fairly well for skilled migrants, since
entrepreneurs do not generally experience
problems in fi nding good candidates from abroad.
However, fi rms are more likely to fi nd unskilled
workers from the pool of already residing migrants
- encouraging illegal migration and over-stays of
unskilled workers (OECD, 2007e). In this regard,
it is important that public agencies reduce the costs
of hiring workers from abroad. Another
particularity of the system is the strong link
between the immigrant and the job held, since the
visa is cancelled as soon as a particular job is over.
The migrant cannot move from fi rm to fi rm to fi nd
a better match, and this potentially limits the
productive output of the migrant and creates
incentives for fi rms to offer lower wages to
immigrants relative to natives (due to their reduced
mobility and thus bargaining power). In order to
avoid this situation, some public supervision of
potential discrimination against foreign workers
may be required.
One alternative system, not currently
implemented in euro area countries, is the point
system. Such a system is in place in Canada,
New Zealand, Australia and Switzerland and is
being discussed in the UK. This system seeks
the admission of mainly long-term immigrants.
Immigrants are screened on the basis of a
set of characteristics that are considered to
make them more likely to assimilate in the
destination country.103 The system has been
found to be successful in attracting higher
percentages of highly skilled migrants relative
to other countries without such a system.104
One important drawback, however, is that the
national government needs to verify ex-ante
the characteristics of the immigrant. This is
especially diffi cult with regard to characteristics
such as education, where there is no good
system for comparing and evaluating the relative
quality of foreign degrees.105 Moreover, such a
system may reduce the fl exibility with which
Notice that family reunifi cation affects the quantity but also the 101
quality of immigration since the skills that are sought with a
particular selection policy do not need to apply to the individuals
who are reunifi ed.
In most cases, governments set additional quotas by region 102
and occupation once they have identifi ed particular shortages
(for example, France and Spain have a “shortage occupation
list” as opposed to a general cap as in Italy). Usually, there are
diffi culties with matching ex-ante and ex-post needs (see OECD,
2007e).
The system awards points to an individual’s characteristics 103
(e.g. education level), with a minimum number of points
required to allow an immigrant to enter the country.
Baker and Benjamín (1994) and Borjas (2001).104
In order to overcome this problem, some countries such as the 105
United Kingdom have incorporated the salary at origin of a person,
depending on the age and educational degree, in their valuation.
This information is expected to capture the different productivity
of the person, although unfortunately it also links the allocation of
visas to the origin country labour market behaviour.
63ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
fi rms are able to hire immigrants in the face of
labour shortages.
4.3.1 POLICIES AFFECTING IMMIGRANTS ONCE IN
A COUNTRY
INTEGRATING LEGAL IMMIGRANTS
A number of studies have found that immigrants’
skills tend to be underutilised upon arrival
(Chiswick and Miller, 2005; see also Box 2),
and during the initial years after arrival, migrants
may face higher unemployment than natives.
This poor initial performance could be due to
language problems, a limited transferability of
their human capital acquired abroad, lack of
familiarity with a host country’s institutions,
or discrimination. Moreover, depending on
the degree of integration, this situation could
perpetuate over time, affecting not only the
migrants themselves but also their descendants.
Countries have encouraged several initiatives to
help better integrate resident immigrants and to
encourage the education of their children.
COURSES DEDICATED TO IMPROVING LANGUAGE
SKILLS AND PRESENTING COUNTRY’S
ADMINISTRATIVE SYSTEM
Much research stresses the importance of
language in the assimilation of immigrants
(Chiswick and Miller, 1995; González, 2005).
Thus, many countries, e.g. Germany and
Austria, encourage immigrant participation in
language programs. In some cases, however,
there are questions regarding the effectiveness
of such courses.106 In a broader initiative,
France created an offi ce for the reception of
foreigners and migration (ANAEM) in 2005 to
facilitate the reception and integration of
foreigners. However, it is still too soon to
evaluate it. Some countries have linked
complete assimilation with the right to reside
permanently and reunify some family members.
For example, since March 2006, in the
Netherlands anyone who wants a long-term
permit must pass a test including a language
exam and questions about Dutch society. This
measure has signifi cantly increased the efforts
required for individuals to be granted
permanent residence in the Netherlands, and
according to the OECD (2007e) it is a factor
underlying a decrease in the willingness of
immigrants to enter the country.
FACILITATING THE EDUCATION AND INTEGRATION
OF IMMIGRANTS’ CHILDREN
Education and language skills of the second
or even third generation of immigrants play a
role in the future success of these individuals
in the labour market. This is especially the
case for children who enter a country at a later
stage of their childhood, since the completion
of an educational level becomes more diffi cult
with age of entry (see National Research
Council, 1995; and Kate, Bachnan and
Morrison, 2001). Moreover, poor economic
conditions and any lack of assimilation of their
parents can also negatively affect children’s
educational performance. Therefore, policies
to monitor an immigrant’s progress at school
and particularly in basic subjects such as
language and mathematics seem of crucial
importance. In order to encourage school
participation, governments have tended to
promote policies such as subsidising some
costs incurred while a child is at school
(such as transportation or food). Some have also
randomised the distribution of foreign students
among different schools in order to increase the
benefi ts of being schooled with native peers
from other socio-economic conditions.
REGULARISATION OF ILLEGAL MIGRANTS
This procedure is usually used as an exception,
but due to large numbers of illegal immigrants
in recent years, it has been recently implemented
in many European countries including Spain,
Portugal, Belgium, Greece, the Netherlands and
France, and has been debated in Germany. The
procedure usually targets certain categories of
foreigners.107 As for the positive effects of such
To improve language skills of permanent immigrants, 106
compulsory courses to learn basic German were introduced in
Austria in 2006. However, the effectiveness of such measures
are subject to heavy debate.
But its effectiveness depends crucially on the amount of red tape 107
involved and the conditions set out for regularisation, since the
compulsory presentation of many documents can affect the number
of participants. For example, the OECD (2007f) has suggested that
the relatively low participation in Greece’s regularization could
have been due to the high amount of paper work required and a
requirement that a certain number of days had been worked.
64ECB
Occasional Paper No 87
June 2008
programs, literature documents an increase in
productivity following a change in the legal
status (Kossoudji and Cobb Clark, 2002; and
Rivera Batiz, 1999).108 A disadvantage of
regularisation programs is that they do not stop
illegal migration and might even encourage it, if
migrants believe that a government is likely to
repeat the process again at a later date.
Summing up, the economic benefi ts that a host
country can derive from immigration depend
on both the number and the characteristics
of migrants that are allowed to enter the
country, and the incentives created by national
institutional frameworks for these immigrants
to fi nd work and to integrate into society rather
than rely on social security or unemployment
benefi ts. Although immigration policy does not
ensure a perfect control over immigration fl ows,
the design of national policies and institutions
are crucial to facilitating the integration of
immigrants and their children into the labour
market and society.
4.4 EDUCATION SYSTEMS AND INCENTIVES 109
The design of the education system, incentives
to invest in education and the amount and
effi ciency of national resources devoted to
education play a key role in encouraging young
people and workers to invest in education
and training, and therefore in determining the
quality of the labour force. Tax and benefi t
systems affect incentives to build up human
capital, and thus the quantity and quality of
education demanded by society.110 Despite
This research fi nds an increase in migrants’ labour opportunities 108
after legalisation, allowing individuals to fi nd a better match for their
characteristics. Moreover, legalisation tends to lead to an increase in
the investment of immigrants in country-specifi c human capital.
Prepared by N. Leiner-Killinger and M. Ward-Warmedinger.109
For example, through subsidies paid for participation in education, 110
tuition fees or taxing persons with high incomes, all of which
differ widely across euro area countries (see Table 18). Generally,
the impact on overall human capital accumulation is diffi cult to
quantify, as it depends crucially on the effi ciency of the resources
spent. A highly progressive tax system associated with high tax
wedges for low-, middle- and high-income groups (here displayed
for both 67% and 167% of the average wage) is usually assumed to
reduce incentives to increase earnings and thus also to improve skills
(via time and fi nancial resources) in human capital accumulation.
Table 18 Selected indicators of tax and benefit systems affecting education incentives
Public subsidies for education in tertiary education 1)
(in % of public expenditure on tertiary education) 2004
Average tuition fees in public institutions (in USD, using PPPs) academic year
2004-2005
Tax wedgeSingle earner 4) 67% of average worker wage
Tax wedge 2)
Single earner 167% of average wage
level 2006
change (p.p) 2001-2006
level 2006
change (p.p) 2001-2006
Belgium 15.7 853 (Fl.) 658 (Fr.) 3) 49.1 -1.6 60.7 -1.6
Germany 17.9 n.a. 47.4 -0.3 53.8 -1.2
Ireland 14.8 no tuition fees 16.3 -1.1 34.2 -1.6
Greece 5.2 n.a. 35.4 0.3 47.7 3.4
Spain 7.8 668 to 935 35.9 0.6 42.6 0.8
France 7.9 160 to 490 44.5 -3.1 53.2 1.2
Italy 16.7 1,017 41.5 -1.2 49.8 0.0
Luxembourg n.a. n.a. 30.6 -0.6 43.5 -1.4
Netherlands 27.0 n.a. 40.6 1.7 46.0 4.5
Austria 18.1 837 43.5 0.6 50.7 0.1
Portugal 5.4 868 31.7 -0.5 41.7 0.3
Finland 16.7 no tuition fees n.a. n.a. 49.9 -2.2
Slovenia n.a. n.a. 38.9 -2.5 n.a. n.a.
Euro area 13.9 38.0 -0.6 47.8 0.2
Denmark 30.3 no tuition fees 39.3 -1.2 49.5 -1.5
Sweden 28.2 no tuition fees 46.0 -1.8 54.6 -0.5
United Kingdom 23.9 n.a. 30.4 2.3 37.6 2.3
Sources: OECD (2006a), “Education at a glance” and OECD (2007h), “Taxing wages 2005-2006”. Notes: Unweighted averages for the euro area. n.a. not available. 1) Public subsidies for education to households and other private entities as a percentage of total public expenditure on tertiary education including student loans, scholarships and other grants to households. 2) See footnote 1 in Table 13. 3) Academic year 2003-04. 4) See also Table 13.
65ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
marginal reductions in tax wedges in some euro
area countries (see Table 18), for the euro area
as a whole, incentives to increase the quality
of labour supply as measured by this indicator
have generally remained unchanged.
As discussed in Chapter 3, investment in the
quality of tertiary education is also important
for innovation and research, to the benefi t of
countries’ growth potential (see also Box 3).
Table 19 presents information on euro area
education expenditure. It shows that all euro
area countries increased their education
expenditure over the last decade, in line with the
increase in educational level of labour supply
described in Chapter 3. While expenditure
per student in non-tertiary education rose by
41% on average in the euro area between 1995
and 2004, expenditure per student in tertiary
education rose less, by 24%, partly due to
expanding student numbers. By 2004, 1.2% of
euro area GDP was spent on tertiary education,
with the vast majority of funds coming from
the public sector, and annual expenditure per
student was highest for tertiary level students.
However, comparison with fi gures for the
United States suggests that the euro area spends
a signifi cantly smaller proportion of annual
GDP on tertiary education, predominantly due
to far fewer funds from the private sector, and
that the level of expenditure per student in the
euro area is generally lower, particularly so at
the tertiary level. Furthermore, it is very
important how effi ciently money is spent 111, and
to what extent education expenditure affects the
public purse or is funded by the private sector.
Recent work by Aghion et al (2007) suggests a
signifi cantly positive link between a university’s
Chart 13 in Annex 3 shows the lack of a clear correlation 111
between aggregate Pisa scores and annual expenditure per
student for euro area countries.
Table 19 Expenditure on education
Change in expenditure on educational insitutions for
all services per student (1995 to 2004) 3
Expenditure on educational insitutions for
tertiary education as a % of GDP in 2004
Annual expenditure per student in euros 2004 4
Primary,secondary and post-secondary Tertiary Public Private
Primary education
All secondary education
All tertiary education including
R&D activities
Primary to tertiary education
Belgium n.a n.a 1.2 0.1 5267 6152 9398 6364
Germany 105 107 1.0 0.1 3927 6015 9726 6192
Ireland 181 126 1.0 0.1 4303 5643 8104 5328
Greece 2 192 151 1.1 n.a 3647 4137 4439 4075
Spain 136 167 0.9 0.3 3940 5318 7443 5237
France n.a n.a 1.2 0.2 4033 6934 8467 6254
Italy 1), 2) 105 130 0.7 0.3 5865 6225 6129 6129
Luxembourg n.a n.a n.a n.a 10681 14187 n.a n.a
Netherlands 136 101 1.0 0.3 4938 5985 10989 6348
Austria n.a 122 0.8 0.8 6087 6476 11140 7780
Portugal 1), 2) 154 98 0.9 0.1 3715 4895 6144 4610
Finland 122 110 1.7 0.1 4429 5906 9925 6189
Euro area 141 124 1.0 0.2 5069 6489 8355 5864
Denmark 121 123 1.8 0.1 6413 7023 12083 7751
Sweden 117 99 1.6 0.2 5928 6380 12871 7210
United
Kingdom 120 93 0.8 0.3 4715 5627 9114 5770
United States 130 132 1.0 1.9 6988 7887 17838 9597
Source: OECD (2007d), “Education at a Glance”. Note: euro area average is unweighted.1) Data on annual expenditure per student refer to public expenditure only. 2) Data on change in expenditure per student refer to public expenditure/institutions only. 3) Index of change between 1995 and 2004 (Expenditure is expressed in 2004 constant prices, defl ated by GDP defl ator; values represent an index with 1995=100). 4) Converted using PPPs for GDP, based on full-time equivalents, converted from US dollars to euro at January 2004 rates.
66ECB
Occasional Paper No 87
June 2008
level of private funding and its autonomy in
spending its budget on research performance.
Some studies (for instance, Hanushek and
Luque, 2003) fi nd little or no evidence of a
positive link between more resources being
allocated to the education system and test
performance. The OECD points to the existence
of relevant ineffi ciencies in public spending on
(secondary) education. In other words,
governments could either provide the same level
of education outputs with less public spending
or, for the existing level of public spending,
increase the education sector’s performance and
effi ciency. It seems that much can already be
achieved if existing public funds are used more
effi ciently and if incentives for private funding
are enhanced.
Chapter 3 has shown that labour market
participation rates for young people are
generally low while they invest in their
education, and that the participation rate of
higher educated individuals is generally high.
This suggests that a lower participation rate of
the younger group does not present a problem,
provided that this is due to time invested in
quality education that both maximises their
chances of direct entry into the labour market
and facilitates a successful future career.112
Alongside the amount and effi ciency of
expenditure on educational systems in Europe,
an improved organisation of the educational
process is of key importance to ensure the best
quality outcome and facilitate the transition
from education to working life. Organisational
aspects include staff-to-student ratios, total
hours of tuition, the organisation of these hours
across semester and the mixture of study and
work experience. Most important in this context
are governance issues and the incentives created
by educational systems for directors, teachers
and professors to invest into the human capital
and skills of students. Wößmann (2007) fi nds
that student performance is better in countries
where private schools increase competition; in
schools that have the freedom to make
autonomous process and personnel decisions,
combined with a system of external exams that
hold schools accountable for their performance;
and in schools where teachers have both the
freedom and incentives to select the best
teaching methods. Furthermore, individual
incentives to participate in higher education,
vocational training and lifelong learning are
infl uenced by the costs of education relative to
returns (see also Box 9), which include the
expected length of time spent in education,
tuition fees and access to fi nance. Also, on the
job training (Box 8) is expected to contribute to
improvements in productivity and thus could
potentially explain differences in developments
across countries. Some reforms to support
vocational training, life long learning and
education have been undertaken in recent
years.113 Although the availability of data on
these characteristics is somewhat limited,
Table 20 shows considerable cross-country
variation in both staff-to-student ratios and
typical graduation dates across the euro area
countries. On average however, tertiary
education classes tend to be somewhat larger
and typical graduation ages higher than
comparable fi gures for the United States.114
Other institutions that matter for the effective
quality of labour supply include wage-setting
institutions (to ensure appropriate private
return) and the fl exibility of labour contracts
(as determinants of the appropriate allocation of
human capital).
The Bologna process (or Bologna accords)
aims to create the European higher education
area by making academic degree standards and
quality assurance standards more comparable
and compatible throughout Europe. Through
Information available in the EU-LFS on student status 112
suggests that the overall (for 15-64 year olds) participation and
employment rates would be 6.3 percentage points higher at
77.0% and 71.0% respectively (instead of the observed 70.7%
and 64.7%) if all inactive students were to work. The effect of
the inactive student population is larger for 15 to 24 year olds
and would increase the participation and employment rates
of this group by 31.4 percentage points to 76.0% and 68.6%
respectively (instead of the observed 44.6% and 37.2%).
For details, see the European Commission’s LABREF database.113
In comparing graduation ages, one should keep in mind that in 114
Germany, compulsory military or civil service generally takes
place after school and increases the age at which individuals
start their studies, and thus also tends to increase graduation
ages by up to one year.
67ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
the implementation of the Bologna process,
higher education systems in European countries
should be organised in such a way that: (a) it
is easy to move from one country to another
(within the European Higher Education Area)
for the purpose of further study or employment
and (b) the attractiveness of European higher
education is increased so many people from
non-European countries also come to study
and/or work in Europe. The European Higher
Education Area aims to provide Europe with a
broad, high-quality and advanced knowledge
base; and promote greater convergence between
the United States and Europe.
In summary, educational systems and the
effi ciency of national resources devoted to
education play a key role in ensuring a smooth
transition from education to working life and
providing the labour force with relevant skills
for the future, for innovation capacity and thus
for overall labour quality within the euro area.
It is important that national education systems
are well funded and effi cient and provide
positive incentives for young people and
workers to invest in education and training,
and for directors, teachers and professors to
invest into the human capital and skills of
students. Introducing elements of competition
and external tests or exams may enhance
such incentives. It is important that such
measures also include incentives to enhance
the education of pupils with lower skills or
fewer talents.
Table 20 Ratio of students to teaching staff in educational institutions and typical graduation ages
Ratio of students to teaching staff in educationalinstitutions (2005) 1)
Typical graduation ages intertiary education
Primary Secondary Tertiary
Tertiaryeducation
programmes of3 to 5 years 2)
Advanced research
programmes 3)
Austria 14.1 10.9 15.3 22 25
Belgium 12.8 9.8 19.6 22-24 25-29
Finland 15.9 13.9 12.5 22-26 29
France 19.4 12.2 17.3 n.a. 25-26
Germany 18.8 15.1 12.2 25 28
Greece 11.1 8.3 30.2 21-22 24-28
Ireland 17.9 15.5 17.4 21-22 27
Italy 10.6 10.7 21.4 22 27-29
Luxembourg n.a. 9.0 n.a. n.a. n.a.
Netherlands 15.9 16.2 n.a. 22-23 25
Portugal 10.8 8.1 13.2 22 n.a.
Spain 14.3 10.6 10.6 20 25-27
Euro area 14.7 11.7 17.0 22.2 26.4
Denmark n.a. n.a. n.a. 22-24 30-34
Sweden 12.2 13.0 8.9 23-25 27-29
United Kingdom 20.7 14.1 18.2 21 24
United States 14.9 15.5 15.7 21 28
Source: OECD (2006a, 2007d), “Education at a glance”. Note: euro area average is unweighted. 1) By level of education, based on full-time equivalents, for Luxembourg public institutions only; the United Kingdom includes only general programmes in upper secondary education. 2) Tertiary-type A (ISCED 5A), data are available by duration of programme.3) ISCED 6.
68ECB
Occasional Paper No 87
June 2008
Box 8
INCENTIVES FOR FIRMS AND WORKERS TO INVEST IN HUMAN CAPITAL ACCUMULATION AND LIFE-LONG
LEARNING 1
The importance of education for productivity and performance cannot be overemphasised. A
discussion of the relationship between the quality of human capital and growth has already
taken place in earlier sections of this report (see, for example, Chapter 2 and Section 3.5). In
those sections, the focus has been on formal education. But human capital improvements take
place not only before individuals start working but also while they are in work in the form, for
example, of continuous vocational training (CVT) and life-long learning (LLL). The average
annual hours spent on CVT in European enterprises (Table A), the percentage of the population
engaged in LLL activities (Table B) and the percentage of secondary level students enrolled
in apprenticeship schemes (Table C) suggest, however, that only a small percentage of the
population receives structured training outside formal education or combines studying with
vocational training. This is true even if one looks at only large fi rms or at younger individuals, as
suggested by the data presented in Tables A and B, despite the fact that the need for continuous
human capital improvement is becoming ever more pressing in view of rapid technological
progress and population ageing. That said, it should be noted that these measures in general,
and specifi cally those on CVT, do not include some more informal types of training such as, e.g.
learning-by-doing or on-the-job training, which may lie outside structured training programmes.
The limited extent to which structured training (which in a general sense also includes
life-long learning activities) takes place is often attributed to market failures; most training
is of a general nature, i.e. transferable across fi rms (at least within the same industry), capital
market imperfections prevent individuals from borrowing against human capital since returns
are uncertain (Becker, 1962), and it is diffi cult to enforce detailed contracts designed to ensure
the quality (and incentive-compatible fi nancing) of training. For example, fi rms have little
incentive to provide general training, since they are faced with the risk that their trained staff
may be poached by competitors who, having not incurred the training costs, are in a position to
offer higher wages. Despite these factors, however, a number of fi rms provide both fi rm-specifi c
and general training (e.g. German fi rms), while a number of individuals also participate in LLL
activities. The signifi cant diversity in the extent of training provided, both among fi rms and,
moreover, across countries (see Table A), is attributed to differences in industrial organisation
and to institutional features introduced to deal with market failures (see, inter alia, Acemoglu and
Pischke, 1998 and 1999; Finegold and Soskice, 1988; and Stevens, 2001) 2. For example, fi rms in
certain countries are encouraged to provide general training through tax breaks and the provision
of subsidies etc., while incentives are also provided to individuals to engage in life-long learning
through, for example, reductions in taxes, subsidised loans to trainees etc. Differences in the
extent of training also depend on the quality of the education system in each country, since
training builds on the skills that have already been provided in schools.
1 Prepared by D. Nicolitsas and H. Stahl.
2 More specifi cally, recently proposed models in the economics literature (see, inter alia, Stevens, 1996; Acemoglu and Pischke, 1998)
suggest that labour market frictions arising from mobility costs between jobs, the nature of skills, or search frictions in the skilled
labour market could explain why some fi rms do provide general training. Mobility costs arise since it takes time to fi nd an alternative
job or because employers providing the general training have an informational advantage over prospective employers regarding the
true productive capabilities of the workers they have trained. This gives these fi rms some monopsony power, which explains their
decision to invest in general training.
69ECB
Occasional Paper No 87
June 2008
4 STRUCTURAL
POLICIES AND
THEIR EFFECT ON
LABOUR SUPPLY
It appears, however, as also suggested by the theory (see Acemoglu and Pischke, 1999) that
one of the most important factors in determining the extent and quality of training provided
is the existence of a monitoring mechanism. In the absence of a detailed contract between the
employer and the employee to ensure that the latter has received the training paid for, it seems
that the state usually fulfi ls this role by introducing a curriculum and setting standards (exams,
monitoring boards). One of the best examples of this kind is the German dual apprenticeship
system, a concise description of which is provided below.
The German dual apprenticeship system
This system (mainly for pupils aged 15 to 19 with lower and middle educational levels) combines
apprenticeships in a company with vocational education at a vocational school. It covers almost
all sectors of the economy; currently about 2/3 of apprentices are working in service sector
fi rms. The training period lasts for between two and three and a half years. For the practical
part, students receive training in a company for three to four days a week, while the theoretical
part, which is both general (language, politics, economics, religious education and sport) and
trade-specifi c, takes place at a vocational school. The quantity and quality of both the theoretical
and the practical parts are strictly regulated, and companies have to teach a broad spectrum of
tasks related to the particular apprenticeship (in other words there is a large amount of general
training involved). Students need a “vocational education contract” (Berufsausbildungsvertrag)
to begin an apprenticeship and are paid during the apprenticeship while fi nal exams are held,
for example, by the Chambers of Industry and Commerce and the Chambers of Handicrafts. In
Table A Hours in Continuous Vocational Training 1) courses per 1,000 hours worked (all enterprises) by size class, 1999 2)
Table B Percentage of all individuals participating in any kind (formal, non-formal or informal) of training activity by age 1)
(percentage, 2003)
CountrySize class (Number of employees) Age
20-49 50-249 250-499 500-999 1,000+ 25-34 35-44 45-54 55-64 Total
Belgium 5 8 9 9 13 Belgium 51 45 41 27 42
Germany 4 5 4 6 6 Germany 50 45 41 32 42
Ireland 8 8 9 24 7 Ireland 51 52 47 42 49
Greece 1 2 4 3 6 Greece 27 19 13 7 17
Spain 4 5 9 9 11 Spain 33 26 20 14 25
France 5 7 9 12 16 France 61 55 51 32 51
Italy 3 4 5 7 10 Italy 57 52 47 35 49
Luxembourg 3) n.a. 5 n.a. 13 n.a. Luxembourg 86 84 79 75 82
Netherlands 7 10 11 12 14 Netherlands 51 44 39 30 42
Austria 4 4 5 5 8 Austria 90 88 87 93 89
Portugal 1 3 5 10 8 Portugal 54 46 39 33 44
Slovenia 2 3 3 7 7 Slovenia 86 83 80 78 82
Finland 9 8 10 12 13 Finland 85 82 76 66 77
Euro area 4.4 5.5 6.9 9.9 9.9 Euro area 60.2 55.5 50.8 43.4 53.2
Denmark 11 14 10 13 15 Denmark 82 83 80 72 80
Sweden 10 8 11 16 14 Sweden 77 74 71 62 71
United Kingdom 6 7 14 17 6 United Kingdom* 44 42 39 23 38
United States n.a. n.a. n.a. n.a. n.a.
Source: Eurostat. Notes: n.a is not available. Euro area averages are unweighted. 1) Apprenticeship schemes are not included. 2) In January 2008, Eurostat data for a later date (2005) were available for only a few countries. From the available data, it appears that the number of hours spent on CVT courses was still low, yet higher than in 1999. 3) The fi gure for fi rms with between 500 and 999 employees refers to fi rms with over 250 employees.
Sources: Eurostat, LFS and ad hoc module on lifelong learning 2003 (see Kailis and Pilos, 2005). Note: 1) Figures refer to the whole population over the age of 25 independent of activity status. * Informal training not included for the UK. Informal training differs from non-formal training because it corresponds to self-learning through, for example, books, computers, learning centres or educational broadcasting.
70ECB
Occasional Paper No 87
June 2008
2002, an estimated 62% of apprentices found a full-time job once they had passed the fi nal exam,
while 32% were unemployed.3
The dual apprenticeship system has a long tradition in Germany. Recently, however, certain
reform needs have been identifi ed. A signifi cant problem is that the supply of vocational
education contracts is persistently lower than demand. Only 25% of fi rms offer apprentice
opportunities. Many fi rms are too small to participate and fi rms face the risk of losing their
investment (including the sometimes relatively generous apprenticeship wages) if apprentices
are not retained at the end of their training period.4 ,5 A lack of quality applicants can also be an
obstacle. Apprenticeship schemes (356 of them) are considered overly specifi c and discussions
are underway to consider reducing them to about 40 broader occupations. Companies would
then have to teach only those skills that are central to a particular apprenticeship. In an effort to
modernise the dual apprenticeship system, the Vocational Training Reform Act was enacted on
1 April 2005.6 This act facilitates the accreditation of qualifi cations acquired outside the dual
apprenticeship system, for example at full-time vocational schools.
3 These employment and unemployment fi gures should be interpreted with great care since some agreements with trade unions explicitly
specify that fi rms must offer to those apprentices who pass the exam at least a temporary job. After two months of having completed
the apprenticeship scheme, male apprentices of German nationality do quit their job, probably in order to do their military or civil
service, although this cannot be identifi ed with the available data.
4 Monthly wages for apprentices range from €200 in transport to €1,000 for apprentices in hotels and restaurants.
5 According to a survey of 2,500 fi rms, total costs per year and per student are €16,435, half of which are labour costs. Subtracting the
gains from the students’ labour input, there remains a loss of €2,400 per year. However, the fi rm can save €5,800 in the opportunity
costs of hiring an external applicant if the student stays at the fi rm after having received his/her apprenticeship.
6 A more detailed presentation of the reform of the German dual apprenticeship system can be found in the site of the Federal Ministry of
Education and Research, “Reform of Vocational Education and Training” (http.//www.bmbf.de/en/1644.php).
Table C Percentage of upper secondary education students enrolled in apprenticeship schemes (combined school and work-related schemes), 2005
BE DE IE GR ES FR IT LU NL AT PT SI FI Euro area DK SE UK US
3.3 45 3.8 n.a. 2.8 11.3 n.a. 13.6 20 32.7 n.a. 3.7 10.5 14.7 47.7 n.a. n.a. n.a.
Source: OECD (2007d), “Education at a glance”, Table C.1.1. p. 277. Notes: n.a is not available. Euro area unweighted average calculated on the basis of available information.
71ECB
Occasional Paper No 87
June 2008
5 MATCHING
LABOUR SUPPLY
AND DEMAND5 MATCHING LABOUR SUPPLY AND
DEMAND 115
A well functioning labour market requires a
good match between labour supply and labour
demand, which should in turn support high
employment rates. In the absence of a shift in
relative demand, an increase in the relative supply
of a particular group of workers, such as the high
skilled, will result in lower wages (i.e. returns) or
higher unemployment for those workers relative
to workers with less education. Generally,
sector specifi c shocks in an environment of
matching ineffi ciency and wage rigidities will
lead to general wage increases in excess of
productivity growth and upward pressure on
prices. An effi cient matching process, combined
with fl exible wages, should therefore reduce the
risk of upward pressures on wages and infl ation
resulting from shocks to relative labour demand
and support employment creation and output,
facilitating adjustment to monetary policy
actions and economic shocks.116
This chapter reviews the evidence on returns
to education and the relative unemployment of
particular groups in the labour market. It aims
to consider the extent of mismatches between
labour demand and labour supply in the euro
area, updating and extending some of the
analysis undertaken by the European Central
Bank (2002). It includes a brief overview of
how labour and product market institutions
affect the matching of labour supply with labour
demand. It concludes with a discussion of the
implications of globalisation and demographic
and technological change for future labour
supply. The main fi ndings of this chapter include
only limited evidence of increasing returns
to education, high rates of unemployment for
low-skilled workers in some countries (e.g.
9.3% overall, but 17.6% in Germany in 2007)
and non-EU nationals (14.7% in the euro area),
persistent labour market shortages of some skill
groups (e.g. those involved in teacher training
and education, engineering and manufacturing,
and health and welfare qualifi cations), and
high rates of youth unemployment in the euro
area (19.3% for 15-19 year olds in 2007).
Furthermore, unemployment rates vary
signifi cantly by geographical region, suggesting
a lack of cross-regional labour mobility and
insuffi cient regional wage differentiation. There
is evidence that euro area countries may have
been underutilising the increases in female
and non-national labour in recent years. The
challenges facing the euro area imply that labour
markets must not only be adequately prepared
with an appropriately enhanced quantity and
quality of labour supply, but must also rise to
the task of ensuring an effi cient match between
its workforce and labour demand.
5.1 RETURNS TO EDUCATION
Returns to education are a function of the
quality of labour (education and skills) and of
the matching of labour demand with labour
supply. Chapter 3 has shown that an increase
in the supply of highly educated workers has
been an important trend in euro area countries
over the last 20 years. In the face of skill-biased
technological change and a rapid increase in
the demand for skilled workers, empirical
studies have shown a signifi cant upward
shift in returns to education in the United
States, whereas studies for other (including
European) countries do not fi nd such a shift 117
(see Box 9). Chart 10 suggests that for the
United States, an increase in total hours worked
by the highly skilled has been accompanied by
an increase in the group’s hourly wage, whereas
in the euro area, the increase in total hours
worked by the highly skilled has, on average, not
been matched by a signifi cant increase in hourly
wage. 118 A possible explanation for this evolution
in hourly wages and in returns to education for
the euro area is that the higher relative supply of
skills was not matched by an increase in relative
demand for skills in Europe. Alternatively,
the quality of education in Europe may not
match that in the United States (as discussed in
Chapter 4), or Europeans may have tended to
Prepared by M. Ward-Warmedinger.115
See, e.g. Christoffel, Kuester and Linzert (2006).116
Ashenfelter et al. (1999).117
Although trends in total hours and hourly wages by skill group 118
vary across euro area countries.
72ECB
Occasional Paper No 87
June 2008
invest more in particular subject areas that are
less rewarded by the market. Other possible
explanations are a lack of wage differentiation or
relative wage fl exibility between different skill
groups, relatively high taxes on higher incomes
or an increase in competitively priced labour
from abroad, which has led to an adjustment
via quantities instead of wages, thus provoking
an increase in the unemployment of low-skilled
relative to high-skilled workers. In a similar
fashion, an increase in the relative supply of
workers with more experience, for example as
the baby-boom cohort reaches prime working
age, and an increase in female labour supply both
have implications for returns and unemployment
rates of these labour input types, depending on
the existing labour market rigidities in those
particular subgroups. The unemployment rates
of different groups of workers are considered
further in the next section.
Chart 10 Hours worked and hourly real wages by educational attainment-based measure of skill
hours high skilled
hours low skilled
real wage high skilled
real wage low skilled
y-axis: 1990 = 100
x-axis: years
Euro area United States
50
70
90
110
130
150
170
50
70
90
110
130
150
170
1990 1992 1994 1996 1998 2000 2002 200470
80
90
100
110
120
130
140
150
70
80
90
100
110
120
130
140
150
1990 1992 1994 1996 1998 2000 20042002
Source: ECB calculations based on EU Klems data.Note: Skill data are derived from national data on educational attainment, with “high skilled” comprising those with university level education and “low skilled” comprising those with primary and/or secondary education (depending on the country). Total economy data, i.e. manufacturing plus services.
Box 9
RETURNS TO EDUCATION IN THE EURO AREA COUNTRIES 1
Markets provide incentives for individuals to invest in skill acquisition through offering higher
wages or earnings for higher levels of education and/or a specialisation in particular fi elds. A
lack of skill acquisition, either to a higher level or within particular fi elds, may refl ect net returns
to education that are too low. Policies that aim to increase the overall skill level, or the supply
of particular skills in an economy, must therefore be confi dent that the supply of skill will be
met by an adequate demand from fi rms, and thus that appropriate rates of return to investment in
education are in place.
1 Prepared by J. Turunen and M.Ward-Warmedinger.
73ECB
Occasional Paper No 87
June 2008
5 MATCHING
LABOUR SUPPLY
AND DEMANDA driving factor in the trend of mean returns to education for many countries has been changing
returns to higher education. A review of available cross-country evidence on so-called
Mincerian returns to higher education, i.e. the premium in earnings for those with a tertiary
(university level) education relative to those with primary education (Budria and Pereira,
2005) or high school education (Acemoglu, 2003), suggests that returns to higher education
appear to have increased only in three out of twelve euro area countries considered since the
1980s. Thus, over the period 1980 to the mid 1990s, at least, there was no general increase in
the returns to education in the euro area, as could have been suggested by a potential increase
in relative demand for highly educated workers owing to skill-biased technological change.
A possible reason for this includes that demand was more or less met by the increase in the
supply of skills described in Chapter 3.
Individuals’ decisions to acquire human capital, in principle, involve a substantial forward-
looking component and a complex calculation weighing the costs of an additional year spent
in education against the expected benefi ts from the investment over an entire lifetime. For
example, if the alternative involves possibly longer periods of future unemployment, higher
(expected) unemployment benefi ts may contribute to reduced investment in education/
human capital. Furthermore, there may be high “opportunity costs” attached to study or work
vis-à-vis leisure for those young people with a high preference for leisure. Mincerian returns
do not take into account the cost of investing in education borne by the individual, and a
useful alternative measure of the returns to education is thus the internal rate of return.2
Although measurement diffi culties suggest that such estimates should be read with caution,
the OECD (2006a) estimates that the private internal rate of return for individuals obtaining a
university-level degree directly following an upper secondary and post secondary non-tertiary
level of education exceeds 8% in a sample of 11 OECD countries in 2003. Rates of return
were estimated to be as high as 16% in Finland and up to 15% in Belgium (see Table 33 in
Annex 3). Furthermore, this study fi nds lifelong learning to be worthwhile, as the estimated
rates of return to a 40-year-old resuming the next level of their higher education on a full-time
basis were above 6.5% in all countries in 2003, signifi cantly higher than the rates of return for
a younger student in some countries.3
National studies also show that the returns to investment in higher education vary by gender and
subject group. For example, evidence for Germany (Lauer and Steiner, 2000) suggests higher
rates of return to education at technical universities. Ammermüller and Weber (2005) fi nd higher
rates of return to law, business and medicine subjects, and lowest rates of return to education,
arts, agriculture and theology. Studies for Germany and Greece fi nd higher returns to education
for women.4 Furthermore, some countries show that within-group income inequality is increasing
for the most educated workers (Luxembourg, Finland, Austria and Portugal). In this context it
should be noted that well-designed education and high skill levels also provide non-pecuniary
returns to benefi ting individuals (e.g. social status, social contacts), as well as positive external
effects to society (e.g. through learning within social groups and integration effects). Although
2 Defi ned as the rate that equates the costs to individuals of attaining the next highest level of education with the present value of an
individual’s lifetime stream of additional earnings associated with the higher level of attainment.
3 One possible explanation for this fi nding is the existence of cohort effects in educational level. Chapter 3 has shown that over the last
two decades, the supply of workers with a university degree has increased signifi cantly. However, for older workers, higher returns to
higher education may refl ect the relatively short supply of degrees within this cohort. Similarly, investment in education that is well
matched to individual career plans and fi rms’ needs may yield higher returns.
4 See, e.g. Ammermüller and Weber (2005), Kanellopoulos et al. (2004), Magoula and Psacharopoulos (1999), Papapetrou (2007).
74ECB
Occasional Paper No 87
June 2008
5.2 RELATIVE UNEMPLOYMENT RATES OF
DIFFERENT GROUPS OF WORKERS
This section presents evidence of mismatch
between labour supply and labour demand 119 in
the euro area and euro area countries, through
the presentation of unemployment rates for
different labour market groups – i.e. a measure
of excess labour supply 120 - and of the range of
group-specifi c unemployment rates.121 This
latter measure describes the difference between
the highest and lowest subgroup unemployment
rates. A greater range implies a larger relative
labour market imbalance. Information on the
persistence of such imbalances is acquired by
considering the range over time. Large ranges
which persist over time may be indicative of
serious labour market imbalances. This section
fi rst briefl y reviews institutional elements which
may contribute to labour market mismatch. It
then considers the extent of labour market
mismatch by: (i) Educational level and subject
area – here mismatch may occur when the level
and/or specialisation of workers’ qualifi cation
do not match those demanded by employers.
This may result in groups of workers
experiencing problems in fi nding or keeping a
job due to under- or overeducation or
inappropriate subject specialisation. (ii) Region
– here mismatch occurs when the workers in
one region experience excess demand for their
services and workers in another are in excess
supply, related to, for example, language or
other barriers which prevent cross-border and
other geographical labour mobility. (iii) Age,
gender and nationality – Chapter 2 has shown
that the largest changes in the composition of
labour supply are driven by these characteristics.
Differences in unemployment rates by age,
gender and nationality may therefore be
informative about both the impact of labour
market rigidities and the extent to which
increases in labour supply over the recent period
are met by demand. Such labour market
mismatches may refl ect both objective economic
reasons (such as adjustment costs, regulations,
skill level, specialisation, work experience) and
other factors (such as discrimination or
prejudices against certain groups).
5.2.1 INSTITUTIONS AFFECTING THE MATCHING
OF LABOUR SUPPLY WITH LABOUR DEMAND
A number of institutions may slow down or
inhibit the matching of labour supply with
labour demand. First, wages that are not
suffi ciently differentiated, for example, by skill
Two broad types of mismatch of the supply and demand for labour 119
can be identifi ed. First, although it may appear that labour supply
is available to feed existing labour demand, it may take time for
unemployed workers to fi nd jobs, given that workers do not have
full information about available positions and fi rms do not have
full information about available labour. Some amount of this
“frictional” labour market mismatch may be unavoidable, even in a
well-functioning labour market, and its size is determined by the
effi ciency of the process of collecting, processing and assessing
the necessary information for both employers and the unemployed,
and by labour market and product market rigidities. Labour market
mismatch can also be caused by “static” mismatch, where the supply
of and demand for labour do not match, for example as a result of
asymmetric information or time-inconsistent expectations by workers
of fi rms’ skill demand and/or a one-off increase in demand or supply
of, e.g. particular skills. Such one-off shocks may arise from business
cycle developments and/or from infl exible adjustment mechanisms,
e.g. through strict EPL or wage rigidity. The less fl exible markets are,
the more persistent mismatches tend to be.
Alternatively, one could consider measures of excess labour 120
demand, such as vacancy rates. However, no reliable data on excess
labour demand exists for all euro area countries. For example,
in Belgium and Luxembourg, vacancy data suffers from, for
instance, double counting, underreporting and/or selective sampling
problems. An alternative measure of labour market imbalances
could be provided by data on the number of fi rms restricted in
their activity due to labour shortages. This data is however not
available for all countries and for all sectors of economic activity.
From the available data, it appears that in the period 2004-2007,
this percentage has been increasing, and the percentage of fi rms
constrained in their activity is higher in services than in industry.
For an alternative measure based on variance of group specifi c 121
unemployment rates, see Lipsey (1960) and European Central
Bank (2002).
not usually included in estimates of the returns to education, such factors are also important for
assessing benefi ts to education.
Returns to education are important in giving incentives for young people to invest in education and
training. To ensure the future effi ciency of education and training programmes, it is furthermore
important that those investing in higher education are well-placed by ability and subject.
75ECB
Occasional Paper No 87
June 2008
5 MATCHING
LABOUR SUPPLY
AND DEMANDor region may contribute to increasing the
mismatch between labour supply and labour
demand (e.g. by not providing the incentives for
capital to shift to the area/regions with high
unemployment or for workers to move to
regions with labour market shortages), thus
increasing the unemployment rates of some skill
groups and in some regions.122 If relative wage
compression is too strong, low-skilled workers
or workers living in low-productivity regions in
particular may remain unemployed. Similarly,
minimum wages which are too high may price
young and lower skilled workers out of the
labour market. This suggests that wages and
labour costs must adjust fl exibly to refl ect local
and sectoral labour market conditions – such as
regional unemployment rates, productivity
growth and workers’ skills.123
Second, non-wage costs that are too high, or fall
too heavily on one type of labour, may contribute
to wage compression and also increase the
labour market mismatch of certain groups. For
example, if taxes fall on employees in the form
of lower wages, this may have negative effects
on the labour supply of particularly low-wage
earners such as the low skilled. High labour
taxes also reduce the possible incentive and re-
allocation effects of a wage change.
Third, overly strict employment protection
legislation (EPL) has been found to increase
the costs to fi rms of adjusting their workforce,
reducing labour market turnover, and can create
a barrier to hiring. A number of studies have
found that overly strict EPL reduces particularly
employment of workers experiencing problems
entering the labour market, such as young
workers and women (Bertola et al 2002,
Jimeno and Rodriguez-Palenzuela 2002,
OECD 2004). This suggests that EPL should
be designed to distort labour turnover to the
lowest extent possible and coordinated with
other policies, such as unemployment benefi t
systems and active labour market policies124.
In addition, creating more individually tailored
and fl exible contracts would be advantageous125
(for example, giving workers the choice to
negotiate more or less employment protection
from fi rms, based on their individual situation
and preferences).
Fourth, the infl exibility of working hours per
employee may reduce the matching of labour
supply and labour demand, in particular for older
workers and women. On the demand side, fi rms
may prefer to adjust hours of work rather than
employment over the business cycle. Similarly,
some workers may prefer fl exible hours or
part-time work in order to combine family and
working life responsibilities (see also Chapter 4,
section 4.2) or to adjust working time with shifts
in leisure preferences or other activities over
their lifetime. Increased working time fl exibility,
which allows the combination of work and family
life and of phased retirement, may facilitate the
matching of labour supply and demand.
Fifth, product market regulation has been found
to restrict employment in the OECD countries
(Nicoletti and Scarpetta, 2005; Nicoletti et al.,
2001).126 Furthermore, the presence of start-up
costs (in particular administrative burdens on
the creation of new companies) may hinder the
growth of some sectors or industries, creating
bottlenecks in the process of matching the
demand and supply of workers in different
sectors of an economy (see Lopez-Garcia,
2003; and Rogerson, 2003). Policies to increase
competition in product markets, such as
reductions in administrative burdens on start-
ups and the removal of statutory barriers to
entry in certain sectors should help to support
employment creation through the creation of
Many studies highlight the insider-outsider theory of the labour 122
market (e.g. Amable et al., 2007; Bertola et al., 2002). According
to this theory, the labour supply of labour market “insiders” are
not so much affected by union density and centralisation and
employment protection. This means that institutions have a more
limited effect on the supply of prime-age males. In turn, they
predominantly affect young workers and females in particular,
since these are not so attached to the labour market (outsiders).
This may include giving fi rms greater scope to adjust wages to 123
these local conditions, for example through the greater use of
opening clauses and ensuring that wage fl oors such as minimum
wages are not set too high to negatively affect employment and/
or are differentiated by region or age.
See also footnote 82 for a discussion of policies to support 124
fl exicurity.
See, for example, Carnoy et al. (1997)125
Product market regulation can affect the labour supply through 126
different channels.
76ECB
Occasional Paper No 87
June 2008
new fi rms, but also by constraining the wage rents
of insiders due to increasing competition.127
Finally, mismatches that result from a defi cit of
certain skills in the labour market may be eased
through higher investment in human capital
acquisition – through increased lifelong learning
and training and educations systems which put a
greater emphasis on identifying future skill needs.
5.2.2 EDUCATIONAL MISMATCH
In all euro area countries, the unemployment
rate decreases signifi cantly with the level of
educational attainment. Table 21 shows that in
2007, the euro area unemployment rate 128 was
9.3% for those with lower secondary education,
compared with 6.4% for those with upper
secondary education and 4.1% for those with
tertiary education. The unemployment rate for
those with lower secondary (tertiary) education
was therefore signifi cantly higher (lower) than
On the one hand, lower barriers to entry for new fi rms and 127
increased real wages due to lower prices may positively affect
activity. On the other hand, employees’ bargaining position may
be weakened by reducing wage rents. Bassanini and Duval (2006)
document a negative effect from a decline in regulation on the
employment of females. Fiori et al. (2007) document a signifi cant
negative effect from product market regulation (PMR) on overall
employment, implying positive effects from deregulation. These
authors also show that labour and product market regulation
interact in two respects. First, PMR has a larger negative impact
on employment when labour markets are also regulated. Second,
rigid PMR has facilitated labour market deregulation, but not vice
versa. Furthermore, wage moderation has a strong yet positive
impact on employment if product market regulation is low.
It should be noted that methodological changes due to, e.g. 128
census revisions, changes in concepts and in a change from the
use of annual to quarterly or more frequent surveys and data,
have resulted in breaks in the LFS survey over time in many
euro area countries, which may affect trends in unemployment
over time for some countries (see Annex 2). However, it should
also be noted that data on unemployment levels, presented in
this section for spring 2007, are well-harmonised.
Table 21 Unemployment rates and mismatch by the level of educational attainment in the euroarea and euro area countries 1)
Country
Unemployment rate in 2007 (%) Educational mismatch
Low Medium High
Weighted average of subgroups
Range 2007
Change in range between highest and lowest rate (p.p) 6)
1993-1995 2) 1996-2001 2002-2007
Belgium 11.7 6.2 3.4 6.5 8.3 2.5 0.8 -2.4 -0.4 2.0 0.3Germany 17.6 8.3 3.6 8.2 13.9 3.2 1.1 -0.2 0.0 5.2 0.9Ireland 6.4 3.5 2.3 3.8 4.1 -3.1 -1.0 -7.8 -1.3 0.0 0.0Greece 7.1 8.0 5.8 7.1 2.3 0.5 0.2 0.2 0.0 -0.8 -0.1Spain 8.4 6.5 4.7 6.7 3.7 -0.2 -0.1 -2.4 -0.4 0.1 0.0France 4) 11.2 6.8 4.8 7.4 6.3 n.a. n.a. -0.4 -0.1 -0.6 -0.1Italy 6.0 3.8 4.2 4.7 2.3 0.3 0.1 1.9 0.3 -1.5 -0.3Luxembourg 4.2 7) 2.4 7) 2.9 7) 3.1 1.8 7) n.a. n.a. n.a. n.a. n.a. n.a.Netherlands 5) 4.1 2.7 1.7 2.6 2.4 n.a. n.a. -3.4 -0.6 1.7 0.3Austria 3) 8.4 3.6 2.4 4.1 6.0 n.a. n.a. 1.2 0.2 1.2 0.2Portugal 8.2 6.7 6.0 7.6 2.2 n.a. n.a. -2.2 7) -0.4 7) 0.8 7) 0.1 7)
Slovenia 5) 7.1 7) 4.4 7) 2.7 4.3 4.4 7) n.a. n.a. 0.4 7) 0.1 7) -1.3 7) -0.2 7)
Finland 3) 8.9 5.9 3.4 5.4 5.5 n.a. n.a. -3.0 -0.5 -1.5 -0.3Euro area 9.3 6.4 4.1 6.6 5.2 1.7 0.6 -1.4 -0.2 0.3 0.1Denmark 4.0 2.4 2.9 3.0 1.6 -0.9 -0.3 -3.6 -0.6 -0.3 0.0Sweden 3) 6.9 4.3 3.5 4.4 3.4 n.a. n.a. -3.2 -0.5 0.2 0.0United Kingdom 6.0 3.6 2.0 3.6 4.0 -1.3 -0.4 -1.9 -0.3 0.2 0.0
Sources: EU-LFS (spring data) and ECB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data8). In bold are the best three performers in terms of (i) a low level and (ii) a low range of unemployment rates. The weighted average of subgroups may not equal the total unemployment rate due to missing data. 1) 25 to 64 years old; the education levels refer to low: lower secondary education and less, medium: upper secondary education, high: tertiary education. 2) Education data start in 1992. 3) Data start in 1995. 4) Data start in 1993. 5) Data start in 1996. 6) The range refers to the difference between the highest and lowest sub-group unemployment rate. Average annual changes are presented in yellow. 7) Based on fi gures smaller than the Eurostat reliability limit. 8) The underlying fi gures for the mismatch indicator (based on the unemployment rate) are unemployment and employment. For every country and the euro area, these fi gures are compared with the respective Eurostat limits. Whenever the numbers are smaller than the publication limit, they are omitted from the published table (these cases are labelled n.a.). Whenever the numbers are larger than the publication limit but below the reliability limit, they are fl agged separately.
77ECB
Occasional Paper No 87
June 2008
5 MATCHING
LABOUR SUPPLY
AND DEMAND
the total unemployment rate of 6.6%, indicating
the much greater unemployment problems
experienced by people without good academic
or vocational qualifi cations and the stronger
demand for employees with a higher level of
education, relative to the supply of such
workers.
The range in unemployment rates across
educational levels was 5.2 percentage points in
2007 for the euro area as a whole, having
narrowed in the generally favourable cyclical
environment of the late 1990s. In recent years,
there have been pronounced differences in
country-specifi c developments. In Germany,
overall unemployment increased rather steeply,
and the gap in unemployment rates by
educational level widened markedly to 13.9
percentage points in 2007 129. In contrast, Ireland,
which experienced a more favourable trend in
total unemployment, saw a decline in the gap
between the unemployment rates of low-
educated and highly educated persons in the
second half of the 1990s.
Unemployment rates also vary signifi cantly
across type of education in the euro area and
euro area countries (see Table 22). Workers
with humanities and services specialisations
typically experience unemployment rates
around or slightly higher than the national total
unemployment rate. On the other hand, workers
with teacher training and education;
engineering and manufacturing; and health and
welfare qualifi cations typically experience
In Germany, from 2005 onwards, persons capable of work 129
applying for basic social security had to register at the
employment offi ce as unemployed and actively seek work. This
might have contributed to the number of low-skilled unemployed
in the German LFS.
Table 22 Unemployment rates and mismatch by type of education in the euro area and euro area countries 1)
Country
Unemployment rate in 2006 (%) Mismatch (range)
General
Teachertraining
andeducation
Humani-ties
languages and art
Socialscience
business,law
Science math
comput-ing
Engineer-ing manu-facturing
Agricul-ture
Health and
welfareSer-vices
Weighted average of sub-
groups 2) 2006
Change (p.p) 5)
2003-2006 3)
Belgium 5.4 2.4 7.8 5.6 4.4 4.6 n.a. 3.4 8.2 7.0 5.8 -1.2 -0.6 Germany 7.3 4.3 7.6 7.4 6.5 10.4 9.1 5.2 9.2 9.8 6.1 -1.9 -0.6 Ireland 4) 2.9 1.6 4.3 1.8 2.9 2.7 0.4 1.5 2.5 3.4 4.0 n.a. n.a. Greece 7.8 4.2 7) 9.2 9.1 9.4 5.2 8.4 7) 9.2 8.1 7.6 5.2 7) -0.6 7) -0.2 7) Spain 6.7 5.1 7.8 6.7 7.2 7) 4.2 4.1 5.6 8.8 7.3 4.6 -0.9 -0.3 France 14.9 n.a. 9.3 7.2 6.1 5.9 4.5 3.7 9.8 8.1 n.a. n.a. n.a. Italy n.a. 4.5 6.5 5.2 6.4 3.3 4.2 2.5 6.0 5.6 n.a. n.a. n.a. Luxembourg 6.3 n.a. n.a. 3.2 n.a. 1.9 n.a. n.a. 4.6 4.0 n.a. n.a. n.a. Netherlands n.a. 2.1 7) 3.8 7) 2.9 3.7 7) 3.7 2.0 2.5 3.3 3.9 3.4 n.a. n.a. Austria 4.4 n.a. 5.0 7) 3.8 n.a. 3.2 n.a. 2.2 7) 4.4 4.1 n.a. n.a. n.a. Portugal n.a. n.a. 8.2 6.0 5.9 6.4 n.a. n.a. n.a. 7.2 n.a. n.a. n.a. Slovenia 8.0 7) n.a. 7.4 7) 5.9 n.a. 4.5 4.8 3.4 7) 4.5 5.3 n.a. n.a. n.a. Finland 8.0 n.a. 8.4 5.3 6.9 7) 5.1 5.5 2.0 6.9 6.3 n.a. n.a. n.a. Euro area 6) 7.1 4.0 7.7 6.5 6.3 7.1 5.7 4.3 8.0 7.5 4.0 -0.4 -0.1 Denmark n.a. n.a. 5.7 3.2 n.a. 2.1 n.a. 2.5 3.5 3.3 n.a. n.a. n.a.Sweden 6.0 2.8 7.7 5.2 n.a. 4.7 4.4 7) 3.0 4.8 5.2 4.8 -2.4 7) -1.2 7)
United
Kingdom n.a. 1.5 3.5 3.0 3.4 2.5 n.a. 1.8 5.1 3.9 3.6 0.9 7) 0.5 7)
Sources: EU-LFS (spring data) and ECB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. Notes: data stem from different LFS sources than in the case of the other breakdowns and are therefore not fully comparable. The detailed education breakdown is only available for 2006.1) 25 to 64 years old. 2) The weighted average of subgroups may not equal the total unemployment rate due to missing data. 3) Starting 2004 for AT, BE, IE, PT, SE, UK.4) Data for Ireland is 2005. 5) the range refers to the difference between the highest and lowest sub-group unemployment rate. Average annual changes are presented in yellow. 6) EA without Ireland. 7) Based on fi gures smaller than the Eurostat reliability limit.
78ECB
Occasional Paper No 87
June 2008
Table 23 Regional mismatch in the euro area
Country1)
Unemployment rate in 2006 (%) Regional mismatch ( range)Change 3) (percentage points)
Min Max Weighted average of subgroups
2006 2) 1984-19954) 1996-2001 2002-2006
Belgium (11) 4.2 17.7 8.3 13.5 4.2 0.3 -4.0 -0.7 4.1 0.8 Germany (41) 5.4 19.7 10.3 14.3 6.1 0.6 6.4 1.1 -5.1 -1.0 Ireland (2) 4.3 4.7 4.4 0.4 -4.5 -0.8 -3.0 -0.5 -1.1 -0.2 Greece (13) 7.2 14.4 9.0 7.2 2.4 0.2 0.6 0.1 -2.5 -0.5 Spain (17) 5) 5.3 13.5 8.6 8.2 3.7 0.4 -6.3 -1.1 -5.8 -1.2 France (22) 6) 6.1 12.9 9.1 6.8 2.4 0.2 -1.4 -0.2 -2.3 -0.5 Italy (21) 2.7 13.6 6.9 11.0 9.0 0.7 2.2 0.4 -13.0 -2.6 Luxembourg (1) 4.7 4.7 4.7 n.a. n.a. n.a. n.a. n.a. n.a. n.a. Netherlands (12) 2.7 5.2 3.9 2.5 -2.4 -0.2 -2.4 -0.4 0.3 0.1 Austria (9) 3.0 8.9 4.8 5.9 n.a. n.a. 0.8 0.1 1.9 0.4 Portugal (7) 3.9 9.5 8.1 5.6 -2.8 -0.3 -2.9 -0.5 1.0 0.2 Slovenia (1) 6.1 6.1 6.1 n.a. n.a. n.a. n.a. n.a. n.a. n.a. Finland (5) 3.5 11.4 7.8 7.9 n.a. n.a. -5.5 -0.9 -4.6 -0.9 Euro area (162) 7) 2.7 19.7 8.4 17.0 2.5 0.3 -6.3 -1.0 -7.3 -1.5 Denmark (1) 4.0 4.0 4.0 n.a. n.a. n.a. n.a. n.a. n.a. n.a.Sweden (8) 6.0 8.5 7.1 2.5 n.a. n.a. -1.2 -0.2 -1.0 -0.2United Kingdom (37) 2.6 9.0 5.4 6.4 -3.5 -0.3 1.8 0.3 0.5 0.1
Sources: EU-LFS (annual, regional data, at NUTS2-level) and NBB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. The weighted average of subgroups may not equal the total unemployment rate due to missing data. In bold are the three best euro area performers who attain (i) a relatively low level and (ii) low range of unemployment rates. Average annual changes are presented in yellow.Notes: 15 to 64 years old.1) Number of regions in 2006 between parentheses.2) No regional mismatch indicator available for Luxembourg, Slovenia and Denmark, as those countries consist of only 1 NUTS2-region. The result for Ireland has to be taken with caution, as this country only consists of 2 NUTS2-regions.3) The developments of the mismatch indicator can be biased by changes in the number of regions considered, or by changes in their territories. This is especially the case in Greece, where the number of regions increased from 9 to 13 in 1988; in Germany, with changes in 1991 (after the reunifi cation), 1996, 2001, 2002 and 2004; and fi nally in Ireland, where in 1998 8 regions were regrouped into 2. As such, the euro area fi gures must also be considered with caution. In the UK, the number of regions increased strongly in 1996.4) 1985-1995 for Germany, 1987-1995 for Spain, Portugal and the euro area, and 1990-1995 for Ireland.5) Ceuta and Melilla are considered to be part of the Andalucia region.6) Excluding the French overseas departments, for which data are only available from 2001 onwards. As those departments have very high unemployment rates, including them would lead to a strong increase of the mismatch indicator over time, which does not provide a correct view.7) The euro area fi gure refl ects the dispersion of unemployment rates within and among the 13 euro area countries. It is not available before 1996.
unemployment rates lower than national total
unemployment rates. These broad patterns also
hold for tertiary level education holders only
and young people aged 15 to 29 years
(see Table 34 and 35 in Annex 3). These latter
specialisations therefore present areas where
the labour demand for skills is high relative to
supply. Furthermore, there are large labour
market imbalances in some subject areas across
countries.130 In general, shortages in the supply
of particular skills may suggest national
education and training systems’ failure to
identify and/or provide the skills demanded by
fi rms. Alternatively, they may indicate that a
subject area is relatively unattractive to workers
– for example due to low rates of return and/or
a lack of relative wage fl exibility.
The broadly stable range in unemployment rates
shows that mismatch by subject specialisation
in the euro area as a whole remained almost
constant over the period considered, again
suggestive of a persistent labour market
misbalance.
For example, many unemployment rates in 2006 are signifi cantly 130
lower than international estimates of the average NAIRU
for the euro area, which tend to cluster around 8% in 2006
(see European Central Bank, 2006).
79ECB
Occasional Paper No 87
June 2008
5 MATCHING
LABOUR SUPPLY
AND DEMAND5.2.3 REGIONAL MISMATCH 131
Table 23 presents information on regional
differences in unemployment rates across the
euro area countries. Mismatch between the
demand and supply of labour across regions can
have a number of origins, including uneven
regional economic development, concentrations
of the resident population and changes in a
country’s industrial structure. However, if
labour is geographically mobile and wages are
fl exible, such mismatches should be short-lived
as unemployed workers move from one region
to another or regional wages adjust to refl ect
local labour market conditions, and investment
fl ows to regions with cost advantages. In this
respect, persistent regional mismatch can
therefore provide some evidence on wage
infl exibility and/or labour immobility. Indeed,
Table 23 shows that unemployment rates
generally vary considerably across regions of
the euro area. Looking at the 2006 levels, there
are relatively large ranges of unemployment
rates in a number of countries. Whilst regional
mismatch appears limited in the Netherlands
and Ireland132, it is very high in Germany,
Belgium and Italy, suggesting a combination of
low regional mobility and low wage fl exibility
to local conditions133. In a number of these
countries clear regional subdivisions appear.
For example, in Belgium134 and Italy,
unemployment is much higher in the southern
regions135; in Germany unemployment is high in
the eastern part of the country and relatively low
in the south.
Regional disparities as measured by the change
in the range of unemployment rates increased
slightly in the euro area over the 1984-1995
period. During the last decade however136,
regional mismatch in the euro area decreased
strongly; in particular, favourable developments
were recorded in Spain, Italy and Finland.
Since 2001, regional dissimilarities also
diminished strongly in Germany and only
clearly increased in Belgium. The decreases in
the regional variation in unemployment rates
are most likely driven by improvements in the
economic environment. However, they may also
provide an indication of some improvements
in the mobility of labour and regional wage
fl exibility, albeit from a low level.
5.2.4 MISMATCH BY AGE, GENDER AND
NATIONALITY 137
Table 24 shows that the unemployment rate of
young people is very high and higher than that of
other age groups in the euro area. In 2007, youth
unemployment (those aged 15 to 24) stood at
almost 15%, nearly three times the unemployment
rate of both prime-aged workers (aged 25
to 54) and those aged 55 to 64. Furthermore,
unemployment among teenagers aged 15 to
19 reached 19.3% and was higher than among
young adults aged 20 to 24.138 High rates of youth
unemployment are of particular concern since
young people should arguably constitute the
group most interested in building up their human
capital and career and who are most fl exible,
both in terms of their subject specialisation and
geographical mobility. Table 36 in Annex 3
shows that rates of unemployment are particularly
high for young people with lower secondary
education and less, across all euro area countries.
But in some countries rates are also very
high for those with upper secondary and even
tertiary education. High youth unemployment
therefore seems to refl ect a major failure of
education systems in identifying and providing
the appropriate level and area of skill demanded
Prepared by J. De Mulder.131
In addition, the average unemployment rate is also low in the 132
Netherlands and Ireland.
See Vamvakidis (2008).133
In Belgium, part of the regional mismatch may also be due to 134
language barriers between the northern and southern regions,
where different languages are spoken.
In Belgium, the regions Brussels and Wallonia experience 135
relatively higher unemployment. In Italy, unemployment rates
are proportionally high in the whole southern part of the country
(Mezzogiorno) and on the islands of Sicily and Sardinia.
The number of regions considered hardly changed over the 136
period 1996-2006, so the recorded developments for the euro
area are more reliable for this period.
Prepared by M. Ward-Warmedinger.137
Although for some countries it should be noted that while the 138
unemployment rate for this group is high, the participation rate
is quite low.
80ECB
Occasional Paper No 87
June 2008
by fi rms. It may also refl ect institutional barriers
that prevent job creation and can cause a (further)
deterioration of human capital, demoralisation
and future labour market diffi culties for the
affected individuals. The pattern of higher rates
of youth unemployment is consistent across all
euro area countries, with particularly high rates
in Belgium, Spain, France, Italy, Luxembourg
and Finland in 2007. Relatively low rates seem
to coincide with countries that have a strong
vocational training system (such as Austria and
Germany – see also Box 8). Unemployment
rates of those aged over 55 were also relatively
high in Germany compared with other euro area
countries.
Whilst the range of unemployment rates
across age groups in the euro area decreased
signifi cantly over the 1983 to 2001 period,
it has increased again since then (mismatch
across the level of education has also increased
for the young, see Table 36 in Annex 3).
It stood at a high level of 12.8 percentage
points in 2007. Overall, this is suggestive of a
persistent and structurally driven labour market
imbalance for the euro area as a whole. The
lack of a positive development in recent years is
worrisome, not least since unemployment rate
differences between age groups are affected by
demographic factors. Following an ageing and
shrinkage of the euro area population and an
associated decrease in the labour supply from
the young (see Chapter 3), these factors might
have been expected to even out the variation in
unemployment rates across age groups to some
extent. In addition, this aggregate fi gure masks
particularly high levels of and rapidly growing
mismatches in some countries in recent years.
Two other important changes in the composition
of labour supply in the euro area over the past
25 years are worth consideration with regard
Table 24 Unemployment rates and mismatch by age groups in the euro area and euro area countries 1)
Country
Unemployment rate in 2007 (%) Mismatch by age (range)Change (p.p) 5)
15-19 20-24 25-54 55-64
Weighted average of subgroups 2007 1984-1995 1996-2001 2002-2007
Belgium 30.2 17.6 6.8 3.8 7.7 26.3 2.9 0.2 -11.2 -1.9 8.2 1.4 Germany 13.5 11.6 7.8 10.1 8.6 5.6 -8.7 -0.7 1.8 0.3 -0.1 0.0 Ireland 2) 13.9 7.4 4.0 2.6 4.6 11.3 -1.3 -0.1 -13.9 6) -2.3 6) 5.2 6) 0.9 6) Greece 24.9 21.5 7.6 3.4 8.2 21.5 8.8 0.7 -0.1 0.0 -9.2 -1.5 Spain 29.1 14.9 6.9 5.6 8.0 23.5 -1.9 -0.2 -13.9 -2.3 -0.5 -0.1 France 28.6 18.6 7.5 6.6 8.7 22.0 -1.3 -0.1 -6.9 -1.2 4.6 0.8 Italy 3) 29.2 16.3 5.0 2.3 5.8 26.9 0.3 0.0 -4.2 -0.7 -3.9 -0.7 Luxembourg n.a. 11.9 6) 3.3 n.a. 3.9 n.a. n.a. n.a. n.a. n.a. n.a. n.a. Netherlands 4) 8.7 4.0 2.5 3.5 3.2 6.2 13.9 -1.2 -7.1 6) -1.2 6) 1.3 6) 0.2 6)
Austria 8.9 7.5 4.2 3.3 4.7 5.7 n.a. n.a 0.2 0.0 2.6 0.4 Portugal 4) 24.8 13.2 7.8 6.8 8.4 17.9 -6.7 6) -0.7 6) -3.2 6) -0.5 6) 9.0 6) 1.5 6) Slovenia 5) 7.4 6) 8.0 4.4 6) 3.1 6) 4.7 5.0 6) n.a. n.a -5.7 6) -0.9 6) -12.8 6) -2.1 6) Finland 33.7 14.6 5.3 5.8 7.8 28.4 n.a. n.a -7.9 -1.3 -3.5 -0.6 Euro area 19.3 13.8 6.6 6.5 7.5 12.8 -7.2 -0.6 -6.9 -1.1 3.1 0.5 Denmark 9.4 5.4 2.7 4.2 3.6 6.7 -9.8 -0.8 0.5 0.1 1.5 0.3 Sweden 6) 37.8 15.3 4.5 4.0 7.0 33.9 n.a. n.a 1.0 0.2 19.6 3.3 United Kingdom 20.7 10.6 3.7 3.3 5.2 17.4 -7.6 -0.6 0.2 0.0 7.2 1.2
Sources: EU-LFS (spring data) and ECB calculations.Notes: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. In bold are the three best euro area performers who attain (i) a relatively low level and (ii) low range of unemployment rates. The weighted average of subgroups may not equal the total unemployment rate owing to missing data. 1) 15 to 64 years old. 2) Data start in 1995. 3) Data start in 1986. 4) Data start in 1996. 5) Average annual changes are presented in yellow. The range refers to the difference between the highest and lowest subgroup unemployment rate. 6) Based on fi gures smaller than the Eurostat reliability limit.
81ECB
Occasional Paper No 87
June 2008
5 MATCHING
LABOUR SUPPLY
AND DEMAND
to mismatch – namely the rapid increase in
the labour participation of women and in
immigration to the euro area (for details of these
developments see Chapter 3).
Table 25 shows that the unemployment rate of
non-nationals was above that of nationals in
2007, particularly so for non-EU nationals, and
across most countries of the euro area.139 The
high rate of unemployment for non-EU nationals
may partly refl ect country-specifi c immigration
experiences with, e.g. asylum seekers and/or the
magnitude of immigrant stocks, and fl ows and
should be compared with the information on the
size of the non-national group presented in
Table 7 and Table 30 in Annex 3. For example,
high rates of unemployment for non-nationals
may refl ect temporary bottlenecks in the
employment of new immigrants, resulting from
increases in immigrant fl ows. However, the
generally relatively high rates of unemployment
for non-nationals may also suggest major
ineffi ciencies in the institutions governing the
integration of non-nationals into the labour
market and/or policies determining immigration
in some countries of the euro area. They could
also highlight the diffi culties that some countries
may face if migration increases, but institutions/
policies do not change. The range of
unemployment rates by nationality stood at a
very high level in many euro area countries,
particularly Belgium, Finland and France in
2007, although trends over the 1996 to 2007
period are suggestive of decreasing labour
market imbalances in the euro area as a whole.
Table 38 in Annex 3 shows rates of unemployment are generally 139
higher for non-nationals across both educational level and age
group. Unfortunately the lack of data prevents a more detailed
breakdown of unemployment rates based on many dimensions
at once.
Table 25 Unemployment rates and mismatch by nationality in the euro area and euro area countries 1)
Country
Unemployment rate in 2007 (%) Mismatch (range)
NationalsOther EU15
Non-EU15
of which 12 NM
of which non-EU27
Weighted average of subgroups 2007
Change (p.p) 7)
1996-2001 2002-2007
Belgium 6.9 9.7 27.1 12.7 30.2 7.7 20.2 -8.0 -1.3 -1.0 -0.2Germany 7.9 9.3 19.2 13.2 20.3 8.6 11.4 -1.7 -0.3 3.7 0.6Ireland 2) 4.4 n.a. n.a. n.a. n.a. 4.4 n.a. n.a. n.a. n.a. n.a.Greece 8.2 n.a. 8.0 8.4 7.9 8.2 n.a. n.a. n.a. n.a. n.a.Spain 7.3 10.7 12.1 11.5 12.3 8.0 4.8 2.9 0.5 -3.0 -0.5France 8.2 7.0 28.0 21.2 8) 24.4 8.3 17.3 -0.5 -0.1 -1.2 -0.2Italy 3) 5.7 n.a. 7.7 7.3 7.8 5.8 n.a. n.a. n.a. n.a. n.a.Luxembourg 3.7 3.3 8) n.a. n.a. n.a. 3.9 n.a. 3.1 8) 0.5 8) n.a. n.a.Netherlands 4) 3.1 n.a. 10.0 n.a. 10.4 3.2 n.a. n.a. n.a. n.a. n.a.Austria 3.9 5.3 12.3 7.2 13.5 4.7 8.4 n.a. n.a. 3.2 0.5Portugal 4) 8.1 n.a. 14.7 n.a. 15.2 8.4 n.a. n.a. n.a. n.a. n.a.Slovenia 5) 4.6 n.a. n.a. n.a. n.a. 4.7 n.a. n.a. n.a. n.a. n.a.Finland 7.6 n.a. 21.0 n.a. 26.4 7.8 13.4 n.a. n.a. n.a. n.a.Euro area 7.0 8.2 14.7 11.0 15.5 7.5 7.7 -0.1 0.0 -1.6 -0.3Denmark 3.4 n.a. 11.9 n.a. 12.2 3.7 8.5 n.a. n.a. n.a. n.a.Sweden 6) 6.7 6.1 19.6 24.5 18.9 7.0 13.5 n.a. n.a. 2.8 0.5United Kingdom 5.0 7.1 8.2 6.0 9.1 5.2 3.2 -1.9 -0.3 -3.5 -0.6
Sources: EU-LFS (spring data) and ECB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. In bold are the three best euro area performers who attain a low level of unemployment rates. The weighted average of subgroups may not equal the total unemployment rate due to missing data. See also Tables A11 and A12 in Appendix 3 for information on unemployment by educational level for non nationals. 1) 15 to 64 years old; the non-national population is separated into non-national EU15 citizens and non-national non-EU15 citizens. For the period 2005-07, this last group is further split into the 12 new member states (which together with the EU15 form the EU27) and non-national non-EU27 citizens. 2) Only data for 1998-2004. 3) Only data for 2005-07. 4) Data start 1999. 5) Data start 2002. 6) Data start 1997. 7) The range refers to the difference between the highest and lowest subgroup unemployment rate. Average annual changes are presented in yellow. 8) Based on fi gures smaller than the Eurostat reliability limit.
82ECB
Occasional Paper No 87
June 2008
Box 10
THE GENDER PAY GAP AND DISCRIMINATION 1
The promotion of a lifecycle approach to work includes increasing female labour market
participation and reducing gender gaps in pay. The Lisbon Strategy for Growth and Jobs stresses
the need to narrow gender pay gaps as a component of providing the appropriate incentives
for workers to enter the labour market and fostering employment, improved work quality and
productivity, and social cohesion. Eurostat’s structural indicators show that the unadjusted
gender pay gap 2 has decreased only very slightly in the euro area, from 15% in 1995 to 14% in
2005. Furthermore, while many countries in the euro area experienced a signifi cant decrease in
gender pay differentials over this period, in a few euro area countries, the gap between average
male and female pay actually increased 3.
Part of the unadjusted gender pay gap in the euro area may be explained by differences in the
average levels of human capital, such as work experience, education and training, or other
characteristics such as the average age of working men and women. Microeconomic studies
have therefore attempted to estimate to what extent observable characteristics (such as age, work
experience and educational level) explain the unadjusted gender pay gap. Weichselbaumer and
Winter-Ebmer (2005) undertake a meta-analysis of 260 of these studies, fi nding that the average
“explained” component of the unadjusted gap stands at about 30%, the remaining 70% being
“unexplained” - which in this literature is often attributed to discrimination.
Reasons for the persistence in the gender wage gap include gender segregation by sector,
occupation and function. It is still the case that girls more frequently study traditionally female
subjects such as languages and crowd into lower paying sectors and occupations, and part-
time employment, which tend to lead to lower paying careers 4. Furthermore, in areas where
technological change results in increases in relative demand and higher relative wages (such as
IT and engineering – see also Box 9), women are typically underrepresented in both universities
and fi rms. The gender pay gap also tends to increase throughout the professional career, with
the male/female wage gap widening with age, experience and rank. This can be explained fi rst
of all by the fact that women tend to interrupt their professional career more often than men, as
they often take more family and household responsibilities than their partners, and second by the
fact that women’s educational levels were lower in the past. Gender differences in promotion
propensities (the “glass ceiling”) and job-to-job mobility also play a role, having a marked
positive infl uence on wage changes and thus on the development of gender wage gaps.
All in all, weaknesses in the labour market situation of women exist that may reduce their incentives
and opportunity for participating in the labour market and ascending the career paths it offers.
Encouraging girls to invest in education and in a greater variety of fi elds are steps in the right direction,
but will not suffi ce unless contemporaneous measures are undertaken to improve equal opportunities
at work. It is necessary to enhance incentives (and reduce rigidities which hinder fi rms) for paying
wages in line with individual productivity; it is also necessary to continue promoting gender equality
1 Prepared by M.Ward-Warmedinger.
2 The unadjusted gender pay gap is given as the difference between average gross hourly earnings of male paid employees and of
female paid employees as a percentage of average gross hourly earnings of male paid employees. The population consists of all paid
employees aged 16-64 that are ‘at work over 15 hours per week’.
3 For more information and data, see Eurostat’s Structural Indicators.
4 See, for example, European Commission (2002b) for a survey of national evidence on these issues.
83ECB
Occasional Paper No 87
June 2008
5 MATCHING
LABOUR SUPPLY
AND DEMAND
Table 26 presents the relative unemployment
rates for men and women in the euro area. This
shows a higher unemployment rate for women
relative to men in most euro area countries in
2007, which suggests that also the increase in
female labour supply has not been fully matched
by labour demand (despite the generally
lower average wages of female workers – see
Box 10), resulting in the ineffi cient use of some
of this increased pool of labour supply in the
euro area. This gap between male and female
employment rates has decreased somewhat in
the euro area as a whole and in many of the
euro area countries over the period 1983 to
2007. While higher unemployment rates for
women and non-nationals may refl ect objective
economic reasons and mismatch due to, e.g.
skill acquisition, and subject specialisation,
they may also be due to other factors
(e.g. discrimination).
5.3 THE FUTURE DEMAND FOR LABOUR
It is impossible to fully anticipate future labour
demand needs, and the problem of matching
labour demand and labour supply will remain.
Nevertheless, some components of future labour
demand that have general implications for the
desired composition of labour supply can be
expected.
in male and female-dominated sectors and access to higher functions (such as managerial positions)
to allow women to progress in their careers. Increased accessibility to childcare should allow women
to more easily combine work and family (see Section 4.2). Finally, in some cases improvements in
the effectiveness of legislation against discrimination may also be necessary.
Table 26 Unemployment rates and mismatch by gender in the euro area and euro area countries 1)
Country
Unemployment rate in 2007 (%) Mismatch (range)Male Female Weighted
average of subgroups 2007
Change (p.p) 6)
1984-1995 1996-2001 2002-2007
Belgium 6.7 8.8 7.7 2.0 -4.8 -0.4 -3.6 -0.6 0.7 0.1 Germany 8.5 8.8 8.6 0.3 -0.7 -0.1 -2.6 -0.4 0.3 0.0 Ireland 4.9 4.3 4.6 0.6 -1.4 -0.1 0.2 0.0 0.3 0.0 Greece 5.0 12.8 8.2 7.7 1.5 0.1 1.5 0.2 -1.4 -0.2 Spain 3) 6.1 10.5 8.0 4.4 7.2 0.6 -5.1 -0.8 -3.3 -0.6 France 8.2 9.3 8.7 1.0 -0.3 0.0 -0.6 -0.1 -2.5 -0.4 Italy 4.6 7.5 5.8 2.8 -1.6 -0.1 -1.4 -0.2 -2.8 -0.5 Luxembourg 4.1 3.5 7) 3.9 0.6 7) -0.5 0.0 -1.7 -0.3 -0.0 7) -0.0 7) Netherlands 2.8 3.7 3.2 0.8 -0.3 0.0 -1.8 -0.3 0.1 0.0 Austria 2) 4.4 5.0 4.7 0.7 n.a. 0.0 -0.8 -0.1 0.5 0.1 Portugal 3) 6.9 10.0 8.4 3.1 -3.7 -0.3 0.9 0.1 0.9 0.1 Slovenia 5) 3.7 5.9 4.7 2.2 n.a. 0.0 0.1 0.0 1.6 0.3 Finland 2) 7.5 8.1 7.8 0.6 n.a. 0.0 -0.9 -0.1 -0.2 0.0 Euro area 6.6 8.6 7.5 1.9 0.3 0.0 -1.6 -0.3 -1.0 -0.2 Denmark 3.3 4.1 3.6 0.8 1.8 0.1 -1.8 -0.3 -0.4 -0.1 Sweden 4) 6.5 7.5 7.0 0.9 n.a. 0.0 -1.4 -0.2 0.2 0.0 United Kingdom 5.5 4.9 5.2 0.7 1.0 0.1 -2.2 -0.4 -0.4 -0.1
Sources: EU-LFS (spring data) and ECB calculations. For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. In bold are the three best euro area performers who attain (i) a relatively low level and (ii) low range of unemployment rates. The weighted average of subgroups may not equal the total unemployment rate due to missing data. 1) 15 to 64 years old. 2) Data start in 1995. 3) Data start in 1986. 4) Data start in 1995.5) Data start in 1996. 6) The range refers to the difference between the highest and lowest subgroup unemployment rate. Average annual changes are presented in yellow. 7) Based on fi gures smaller than the Eurostat reliability limit.
84ECB
Occasional Paper No 87
June 2008
First, an increasing demand for higher skills
resulting from technological developments can
be expected. While this evolution could help to
ease some of the tensions in countries or regions
with an over-education 140 problem, ceteris paribus it will aggravate the existing mismatches
in others. Furthermore, a change in the
production structure of the economy owing to
technological progress may lead to relative
changes in demand for educational groups.
Second, the consequences of an ageing
population and the rising proportion of the
elderly will affect both the size and the
composition of the euro area labour force. The
shrinking of the total labour force and the ageing
of society will increase demand for certain
subject-specifi c skills such as medical workers
and caregivers in the euro area.
Third, the ongoing process of globalisation may
infl uence the structure of future labour demand.
Changes in the structure of international trade
in goods and services may lead to changes in
labour demand. For example, increased trade
with countries with a large supply of relatively
low-wage, low-skilled labour is likely to induce
a decrease in demand for low-skilled workers
in the production of tradable goods in the euro
area. Similarly, this may increase the demand
for high-skilled workers in the euro area as
emerging markets seek to import capital goods
and technology from abroad. The effects of
demographic change and the increased demand
for services may mean that the demand for
labour in services, including caregivers and
medical workers, will further increase.
These factors mean that the euro area labour
markets must not only be adequately prepared
with an appropriately enhanced quantity
and quality of labour supply, but must
also rise to the task of ensuring a suffi cient
match between this workforce and labour
demand. This may imply, for instance,
greater investment by national authorities in
education systems and greater incentives for
schools, universities and fi rms to identify and
cultivate those skills sought by fi rms. More
general competences will also be needed to
allow individuals/students to adapt fl exibly
to new developments in labour demand.
Globalisation presents a risk to low-skilled
people by increasing the supply of foreign
goods produced using lower wage low-skilled
labour abroad. It is important that wages are
fl exible – to provide clear signals with regard
to which skills are demanded by fi rms and to
reduce wage rents, which privilege ‘insiders’
thereby supporting employment creation.
Furthermore, regulations and institutions
should be reformed to reduce the costs and
ineffi ciencies of the labour marketing process.
Lastly, more coordinated efforts across the
euro area to eliminate the remaining barriers
to geographical mobility (languages, transfer
of pensions, etc.) are needed.
Over-education refers to a situation where workers are undertaking 140
tasks of a lower level than that for which they are qualifi ed.
85ECB
Occasional Paper No 87
June 2008
ANNEX 1
ANNEX 1 MEASUREMENT ISSUES141
When addressing labour supply, it has to be noted
that offi cial statistics may potentially mis-measure
the true employment and labour market
participation rates. This annex briefl y reviews the
main measurement issues relating to accurately
capturing labour supply. First, the labour force is
equal to the sum of the population at work and
unemployed job-seekers. In the LFS, the ILO-
defi nitions of “at work” 142 and “unemployed” are
used (see also Annex 2). However, some
individuals who do not work but are available for
employment, and thus are a part of actual labour
supply, are not captured by the standard defi nition
of unemployment, which results in an
underestimation of actual labour supply. Box 11
discusses the ILO defi nition of unemployment and
considers how this affects measured employment.
Second, some work is not captured by statistics
on employment because individuals work in the
unoffi cial (or “shadow”) economy. This results
in an underestimate of employment, and to the
extent that these workers are measured as inactive,
an underestimate of actual labour supply, when
supply is measured by the participation rate.
Third, some services that could be provided
through the market or by government are
produced in the household instead. Different
propensities for household production across
countries or over time may infl uence measures
of labour supply. In particular there has been
an increasing trend towards market-based or
government provision of services that have been
traditionally produced within the household (e.g.
childcare). This trend has affected female labour
supply in particular. Freeman and Schettkat
(2005) collect and assess information on the
amount of hours worked in the household across
countries based on time-diaries. They conclude,
for example that there is a more signifi cant
provision of services typically considered as a
part of household production through the market
in the United States than in European countries.
The extent to which household production is
provided via the market is also likely to depend on
the supply of suitable, typically low-skilled and
low-wage labour. Furthermore, part of services
that are “outsourced” from the household may
rather fall into the shadow economy. Generally
as female labour supply increases these effects
may become more important.
Prepared by J. Turunen and M.Ward-Warmedinger.141
Persons who during the reference period performed work for a 142
wage or salary, or for profi t or family gain, in cash or in kind,
for at least one hour. This includes those with a job or enterprise
who are not at work due to temporary absence.
Box 11
ILO DEFINITION OF UNEMPLOYMENT 1
Information on labour market status is computed from data collected by national labour force surveys
in many developed countries. Labour force statistics generally divide the adult population into three
mutually exclusive groups: the employed, the unemployed, and the inactive (i.e. people out of the
labour force). The European Community Labour Force Survey is based on the ILO-defi nition of
“at work” and “unemployed”. Under this defi nition, the employed comprise all persons aged 15
and over 2 who, during some reference period, were engaged in paid employment (including those
working in a family business). The LFS additionally specifi es that individuals are employed if they
engage in paid employment for at least one hour per week. This includes those with a job or enterprise
who are not at work due to temporary absence. In the LFS, people between the ages of 15 and 74 3
are classifi ed as “unemployed” if they meet all of the following requirements: (1) they are without
1 Pepared by E. Viviano.
2 16 and over in Spain, the United Kingdom and Sweden (1995-2001); 15-74 years in Denmark and Sweden (from 2001 onwards).
3 Those aged 16-74 in Spain, Sweden (1995-2000) and the United Kingdom.
86ECB
Occasional Paper No 87
June 2008
work (work less than one hour per week); (2) they state that they are seeking employment for at
least one hour per week; (3) they are available to start work within the following two weeks; (4) they
sought employment at some time during the previous four weeks (see Eurostat 1996). People neither
employed nor unemployed are considered inactive (and are excluded from the labour force).
People out of the labour force are thus a composite group formed by persons who do not want a job,
persons who are not searching but might take up a job if offered, and persons who are searching for
a job but took their last step to search for one more than four weeks ago. Brandolini et al. (2006)
calculate that, on average in European countries, about a fi fth of all people who declared they were
seeking work in the 1990s were left out of the labour force on the basis of this four-week requirement.
Because of the sheer size of this group – also labelled “potential labour force” or simply “potentials” –
Brandolini et al. investigate the role of the four-week criterion in determining the size of unemployment
and conversely of the potential labour force. They test whether transition probabilities differ among
the unemployed, the potentials, and the other inactive persons in European countries and fi nd that
the (annual) transition probabilities of potentials are always different from those observed for other
people out of the labour force, whereas in some cases they can be considered similar to those of
the unemployed. On this basis, they conclude that in the European labour markets a sizeable “grey
area” exists between the states of unemployment and inactivity. The European labour markets would
be better described by four distinct states (employed, unemployed, potentials, and other inactive
population) than by the three-way characterisation of the ILO guidelines, confi rming the conclusion
reached by Jones and Riddell (1999) for Canada and by Centeno and Fernandes (2004) for Portugal.
The ILO four-week requirement can be viewed as a minimum level of search intensity that job seekers
must show in order to be classifi ed as unemployed: at least one search action – such as sending an
application to a potential employer, visiting an employment agency, or (in Europe) simply looking at
newspaper advertisements – must be undertaken in a four-week period.4 However, this condition may
be exceedingly rigid. From the theoretical standpoint the total effort put into a job search depends on
individual resources, search costs, and expected returns; moreover, it is endogenously determined,
given the labour market conditions. As a consequence, whether this arbitrarily set minimum level
of search intensity is a good criterion for distinguishing between active and less-active job seekers is
ultimately an empirical question. Brandolini et al. (2006) identify the search intensity that separates
the unemployed from the potentials by looking at the Italian data. They proxy search intensity by
the “number of months since the last search action” using data from the Italian labour force survey,
the only EU survey where this information is available. They compare the (quarterly) transition
probabilities of the unemployed with those of a group comprising the most-intensive job seekers
among the potentials. They fi nd that the potentials turn out to be behaviourally indistinguishable
from the unemployed when their last search action occurred “not long before” the ILO four weeks
(up to 11 months for certain groups in the population). Letting the boundary between unemployment
and potential labour force be determined by the data, rather than by the arbitrary four-week criterion,
would raise the Italian unemployment rate in 2000 from 11% to 13%. Four weeks may be too short a
period to identify the unemployed, especially in some demographic groups, like people living in the
Southern part of Italy and older women.
The same exercise, repeated for the years 2004, 2005 and 2006, suggests that, by letting the size
of unemployment be determined by the data, would increase the Italian unemployment rate by
1.5 percentage points (see Bank of Italy, 2006).
4 Alternatively, search intensity may be identifi ed with the probability of applying for a job during a given period or with the number of
applications sent per unit of time (as in Petrongolo and Pissarides, 2001).
87ECB
Occasional Paper No 87
June 2008
ANNEX 2
ANNEX 2 DATA SOURCES AND DESCRIPTIONS
A2.1 THE LABOUR FORCE SURVEY:
Euro area data presented and used in this paper
are drawn from the European Community
Labour Force Survey, which has been conducted
since 1983. The annual data used consist of the
spring surveys (quarter 1 surveys for France and
Austria, quarter 2 for all other countries) up to
the fi rst quarter of 2007. Eurostat compiles these
data, and a detailed description of the sampling
methods and adjustment procedures can be found
in “The European Union Labour Force Survey
– Methods and Defi nitions, 2001” (http://circa.
europa.eu/irc/dsis/employment/info/data/eu_lfs/
index.htm). The available variables are listed
and described in the “EU Labour Force Survey
database – User guide”. The need to preserve an
international comparability of the data means
that the LFS dataset uses standardised and
widely accepted defi nitions of, e.g. employment
and unemployment, as adopted by the ILO.
These constitute the basis of the Eurostat LFS.
It should be noted that these defi nitions differ
from those adopted by countries in their national
defi nitions of labour market status, where
international comparability is not a necessity.
No data are available for Spain and Portugal
prior to 1986, for Austria and Finland prior to
1995 and for Slovenia prior to 1996.
The LFS is based on the ILO-defi nition of “at
work” and “unemployed”. In the LFS, persons
aged 15 and over 143 are defi ned as “at work” or
employed if during the reference period they
performed work for a wage or salary, for profi t
or family gain (e.g. in a family business), in cash
or in kind, for at least one hour. 144 This includes
those with a job or enterprise who are not at
work due to temporary absence. People aged 15
to 74 145 are classifi ed as “unemployed” if they
meet all of the following requirements: (1) they
are without work (work less than one hour per
week); (2) they state that they are seeking
employment for at least one hour per week; (3)
they are available to start work within the
following two weeks; (4) they sought
employment at some time during the previous
four weeks. People neither employed nor
unemployed are considered inactive (and are
excluded from the labour force). According to
the LFS defi nition, “inactive people” are a group
formed by persons who do not want a job,
persons who are not searching but might take a
job if offered, and persons who are searching
for a job but took their last step to fi nd one more
than four weeks before the interview. Box 11
discusses the implications of these defi nitions
further.
Nationality is interpreted as citizenship.
Citizenship is defi ned according to national
legislation of each country. Education level is
classifi ed according to the International Standard
Classifi cation of Education 1997 (see below).
The expression ‘level successfully completed’
is associated with obtaining a certifi cate or a
diploma, when there is a certifi cation. In cases
where there is no certifi cation, successful
completion must be associated with full
attendance. When determining the highest level,
both general and vocational education/training
is taken into consideration.
Data on the hours of work measure the number of
hours usually worked per week. This covers all
hours including extra hours, either paid or unpaid,
which the person normally works, but excludes
travel time between home and workplace and
time taken for the main meal break (usually at
lunchtime). Persons who usually work from
home are asked to include the number of hours
they usually work there. Apprentices, trainees and
other persons learning a job are asked to exclude
any time spent at college or in other special
training centres. Some persons, particularly self-
employed persons and family workers, may not
have usual hours, in the sense that their hours
vary considerably from week to week or month
to month. If a respondent is unable to provide a
fi gure for usual working hours for this reason, the
average of hours actually worked per week over
16 and over in Spain, the United Kingdom and Sweden (1995-143
2001); 15-74 years in Denmark and Sweden (from 2001 onwards).
This group therefore includes employees, employers employing 144
one or more employees, the self-employed and family workers.
Those aged 16-74 in Spain, Sweden (1995-2000) and the United 145
Kingdom.
88ECB
Occasional Paper No 87
June 2008
the past four weeks is used as a measure of usual
hours. The distinction between full-time and part-
time work is based on a spontaneous response
by the respondent (except in the Netherlands,
where part-time is determined if the usual hours
are fewer than 35 hours and full-time if the usual
hours are 35 hours or more, and in Sweden where
this criterion is applied to the self-employed).
It has to be noted that methodological changes
due to, e.g. census revisions, changes in concepts
and a change from the use of annual to quarterly
or more frequent surveys and data, have resulted
in breaks in the LFS survey in many euro area
countries. Details of these breaks can be found
at: http://circa.europa.eu/irc/dsis/employment/
info/data/eu_lfs/F_LFS_COMPARABILITY.
htm. Comparability of developments over time
in this report is optimised through the use of the
spring surveys for all countries over the relevant
time period considered.
DATA EXTRAPOLATION
In the case of Germany, data prior to 1991
have been obtained on the basis of the
developments in West Germany. The euro
area aggregate refers to the 13 countries that
formed the euro area in 2007. Data prior to
1996 have been obtained on the basis of the
growth rate of the largest aggregate available
(i.e. 13 countries since 1996, 12 countries in
1995, 10 countries from 1986 to 1994 and
8 countries for 1983 to 1985).
DEFINITION OF EDUCATIONAL LEVELS IN THE
EU-LFS
In the EU-LFS, the ISCED 1997 classifi cation
is used. Respondents are considered as low,
medium or high skilled if their most advanced
diploma is respectively ISCED 0, 1 or 2;
ISCED 3 or 4; and ISCED 5 or 6.
ISCED 1 - PRE-PRIMARY EDUCATION
Programs at level 0, (pre-primary) defi ned as
the initial stage of organised instruction are
designed primarily to introduce very young
children to a school-type environment, i.e. to
provide a bridge between the home and a school-
based atmosphere. Upon completion of these
programs, children continue their education at
level 1 (primary education).
ISCED 2 - PRIMARY EDUCATION OR FIRST STAGE
OF BASIC EDUCATION
Programmes at level 1 are normally designed on a
unit or project basis to give students a sound basic
education in reading, writing and mathematics
along with an elementary understanding of other
subjects such as history, geography, natural
science, social science, art and music. In some
cases religious instruction is featured. The core
at this level consists of education provided for
children, the customary or legal age of entrance
being not younger than fi ve years or older than
seven years. This level covers, in principle, six
years of full-time schooling.
ISCED 3 - LOWER SECONDARY EDUCATION OR
SECOND STAGE OF BASIC EDUCATION
The contents of education at this stage are
typically designed to complete the provision of
basic education which began at ISCED level 1.
In many, if not most countries, the educational
aim is to lay the foundation for lifelong learning
and human development. The programmes at
this level are usually on a more subject-oriented
pattern using more specialised teachers and more
often several teachers conducting classes in their
fi eld of specialisation. The full implementation
of basic skills occurs at this level. The end
of this level often coincides with the end of
compulsory schooling, where it exists.
ISCED 4 - (UPPER) SECONDARY EDUCATION
This level of education typically begins at the
end of full-time compulsory education for those
countries that have a system of compulsory
education. More specialisation may be observed
at this level than at ISCED level 2 and often
teachers need to be more qualifi ed or specialised
than for ISCED level 2. The entrance age to this
level is typically 15 to 16 years. The educational
programmes included at this level typically
require the completion of some 9 years of full-
time education (since the beginning of level 1)
89ECB
Occasional Paper No 87
June 2008
ANNEX 2
for admission, or a combination of education
and vocational or technical experience.
ISCED 3A: Programmes designed to provide
direct access to ISCED 5A;
ISCED 3B: Programmes designed to provide
direct access to ISCED 5B;
ISCED 3C: Programmes not designed to lead to
ISCED 5A or 5B.
ISCED 5 - POST-SECONDARY NON TERTIARY
EDUCATION
ISCED 4 captures programmes that straddle
the boundary between upper secondary and
post-secondary education from an international
point of view, even though they might clearly
be considered as upper secondary or post-
secondary programmes in a national context.
These programmes can, considering their
content, not be regarded as tertiary programmes.
They are often not signifi cantly more advanced
than programmes at ISCED 3 but they serve to
broaden the knowledge of participants who have
already completed a programme at level 3.
Typical examples are programmes designed
to prepare students for studies at level 5 who,
although they have completed ISCED level 3,
did not follow a curriculum which would allow
entry to level 5, i.e. pre-degree foundation
courses or short vocational programmes. Second
cycle programmes can be included as well.
ISCED 4A: See text for ISCED 3
ISCED 4B: See text for ISCED 3
ISCED 4C: See text for ISCED 3
LEVEL 6 - FIRST STAGE OF TERTIARY EDUCATION
(NOT LEADING DIRECTLY TO AN ADVANCED
RESEARCH QUALIFICATION)
This level consists of tertiary programmes
with a more advanced educational content than
those offered at levels 3 and 4. Entry to these
programmes normally requires the successful
completion of ISCED level 3A or 3B or a
similar qualifi cation at ISCED level 4A. They
do not confer an advanced research qualifi cation
(ISCED 6). These programmes must have a
cumulative duration of at least two years.
ISCED 5A: Programmes that are largely
theoretically based and are intended to provide
suffi cient qualifi cations for gaining entry into
advanced research programmes and professions
with high skills requirements.
ISCED 5B: Programmes that are practically
oriented/occupationally specifi c and are mainly
designed for participants to acquire the practical
skills and know-how needed for employment
in a particular occupation, trade or class of
occupations or trades, the successful completion
of which usually provides the participants with
a labour-market relevant qualifi cation
ISCED 7 - SECOND STAGE OF TERTIARY
EDUCATION (LEADING TO AN ADVANCED
RESEARCH QUALIFICATION)
This level is reserved for tertiary programmes
which confer an advanced research qualifi cation.
The programmes are therefore devoted to
advanced study and original research and not
based on course-work only. They typically
require the submission of a thesis or dissertation
of publishable quality which is the product of
original research and represents a signifi cant
contribution to knowledge. They prepare
graduates for faculty posts in institutions
offering ISCED 5A programmes, as well as
research posts in government, industry, etc.
A2.2 US DATA
Total population data stems from the AMECO
database. Working age population, employment
and unemployment are from the US Bureau
of Labour Statistics (BLS). Each month, the
BLS analyses and publishes statistics on the
labour force, employment, and unemployment,
classifi ed by a variety of demographic, social,
90ECB
Occasional Paper No 87
June 2008
and economic characteristics. These statistics
are derived from the Current Population Survey
(CPS), which is conducted by the Census
Bureau for the BLS. This monthly survey of the
population uses a sample of households that is
designed to represent the civilian noninstitutional
population of the United States.
The data used in Table 17 are based on
the publication Social Security Programs
Throughout the World: Europe, 2006. This
information is compiled by the US social security
administration and the International Social
Security Association (ISSA). The publication is
available through: http://www.ssa.gov/policy/
docs/progdesc/ssptw/2006-2007/europe/index.
html. This publication contains an overview
of the different social security programs in the
world, including parental leave structures.
A2.3 PISA
The OECD’s Programme for International
Student Assessment (PISA) is an internationally
standardised assessment that was jointly
developed by participating countries and
administered to 15-year-olds in schools. The
assessment conducted by PISA in 2006 covers the
domains of reading, mathematical and scientifi c
literacy, with the major focus on scientifi c
literacy. Paper-and-pencil tests are used, with
assessments lasting a total of two hours for each
student. Test items are a mixture of multiple-
choice items and questions requiring students
to construct their own responses. Students
also respond to a background questionnaire,
and additional supporting information is
gathered from the school authorities. Fifty-six
countries and regions, including all 30 OECD
member countries, took part in the PISA 2006
assessment. The assessment takes place every
three years.
A2.4 OECD DATA ON YEARLY HOURS WORKED
The OECD uses information from labour force
surveys: for European countries this is the
EU-LFS.
Annual hours are estimated by annualising usual
weekly hours and then by adding or subtracting
all “unusual” events that occurred during the
year, such as additional overtime work, public
holidays, sick, maternity, paid and other leave.
The number of weeks corresponding to each
of the events can be estimated by using the
differences between actual and usual weekly
hours reported in the survey and extrapolating
them over the year. However, for events that
are not randomly distributed over the year, such
as paid leave and public holidays, statutory
information is used. For the European countries,
information concerning collectively agreed paid
leave and public holidays are taken from the
Working Time Developments of the European
Industrial Relations Observatory (EIRO).
The estimation of average annual hours worked
is done separately for full-time employees, part-
time employees and self-employed persons.
Average annual hours for total employment are
then calculated as a weighted average of the
three categories.
91ECB
Occasional Paper No 87
June 2008
ANNEX 3
ANNEX 3 DETAILED TABLES AND CHARTS
Chart 11 Employment rates in the euro area and the United States by gender
(percentages)
Euro area
United States
Males Females
55
65
75
85
55
65
75
85
20041983 1986 1989 1992 1995 1998 200140
50
60
70
40
50
60
70
20041983 1986 1989 1992 1995 1998 2001
Sources: Eurostat and ECB calculations.Note: 15 to 64 year olds.
Table 27 Participation and employment rates for five-year age groups in the euro area
Males Females1983 1995 2001 2007 1983 1995 2001 2007
Participation rates Total 80.4 76.4 77.1 78.4 46.9 54.3 57.9 63.2
15-19 37.9 24.8 26.4 25.4 31.6 19.5 20.9 20.4
20-24 80.3 67.6 68.0 68.1 66.0 59.8 58.2 59.7
25-29 93.2 88.6 88.1 88.7 63.8 72.0 73.9 77.4
30-34 97.5 95.0 94.7 94.4 58.9 70.1 73.6 77.9
35-39 97.7 96.0 95.6 95.5 55.5 69.6 73.3 77.7
40-44 97.2 95.6 95.1 94.8 51.9 69.3 73.6 78.1
45-49 95.5 93.6 93.5 93.3 49.4 62.9 70.4 76.4
50-54 90.6 87.2 87.8 89.9 41.8 53.9 60.2 69.7
55-59 73.6 66.1 66.7 71.6 30.4 35.8 41.0 51.7
60-64 40.3 28.2 30.0 36.3 13.2 11.3 13.3 20.5
Employment ratesTotal 73.5 69.1 71.8 73.2 40.9 46.6 52.3 57.8
15-19 28.3 19.3 22.4 20.8 21.0 13.9 16.9 16.1
20-24 64.9 53.8 58.8 59.3 51.5 45.1 48.8 50.8
25-29 83.9 77.8 80.3 81.3 54.5 59.8 65.1 69.7
30-34 91.3 86.9 88.8 88.7 53.3 60.5 66.4 71.3
35-39 93.0 89.5 90.7 90.7 51.0 61.3 67.1 71.7
40-44 92.8 89.3 90.6 90.0 48.3 62.2 67.8 72.8
45-49 91.2 87.6 89.0 88.8 46.6 57.1 65.5 71.3
50-54 86.1 81.5 83.3 85.2 39.3 48.7 55.8 64.8
55-59 69.7 60.2 61.6 67.0 28.7 32.1 37.1 47.9
60-64 38.0 26.7 28.2 34.2 12.8 10.9 12.7 19.3
Sources: EU-LFS (spring data), ECB and NBB calculations.
92ECB
Occasional Paper No 87
June 2008
Table 28 Participation rates (and employment rates in brackets) by country and subgroup
a) Subdivision according to gender
Average annual change (percentage points) Level (%) Trend developments Recent developments
1984 – 1995 1) 1996 – 2007 2) 1996 – 2001 3) 2002 – 2007 2007
MalesBelgium -0.3(-0.2) 0.1(0.1) 0.1(0.3) 0.1(-0.1) 73.2(68.2)
Germany -0.2(-0.1) 0.1(0.0) -0.1(-0.2) 0.4(0.3) 81.4(74.4)
Ireland -0.6(-0.4) 0.4(0.9) 0.5(1.6) 0.3(0.2) 81.2(77.2)
Greece -0.4(-0.4) 0.1(0.2) 0.0(-0.1) 0.3(0.5) 78.9(74.9)
Spain -0.4(-0.1) 0.5(1.2) 0.5(1.8) 0.6(0.7) 81.6(76.6)
France -0.5(-0.7) 0.0(0.1) 0.0(0.4) -0.1(-0.3) 74.3(68.2)
Italy -0.5(-0.7) 0.1(0.4) 0.1(0.3) 0.1(0.5) 74.5(71.1)
Luxembourg -0.4(-0.4) 0.0(-0.2) 0.0(0.1) -0.1(-0.4) 75.5(72.4)
Netherlands 0.2(0.5) 0.4(0.6) 0.7(1.3) 0.1(-0.1) 84.7(82.3)
Austria n.a.(n.a.) 0.0(-0.1) -0.3(-0.3) 0.2(0.1) 80.2(76.7)
Portugal -0.7(-0.6) 0.2(0.2) 0.5(0.9) 0.0(-0.6) 79.0(73.6)
Slovenia n.a.(n.a.) 0.4(0.6) 0.3(0.5) 0.6(0.8) 76.0(73.2)
Finland n.a.(n.a.) 0.4(1.0) 0.8(1.7) 0.0(0.3) 79.3(73.4)
Euro area -0.3(-0.4) 0.2(0.3) 0.1(0.4) 0.2(0.2) 78.4(73.2)
Denmark 0.1(0.4) -0.1(0.0) -0.4(-0.1) 0.1(0.2) 84.0(81.3)
Sweden n.a.(n.a.) 0.2(0.4) 0.1(0.8) 0.3(0.1) 82.0(76.7)
United Kingdom -0.1(0.0) -0.1(0.2) -0.2(0.5) -0.1(-0.1) 81.6(77.1)
FemalesBelgium 0.6(0.8) 0.7(0.8) 0.5(0.9) 0.9(0.7) 60.2(54.9)
Germany 0.7(0.7) 0.7(0.7) 0.4(0.6) 1.0(0.8) 69.8(63.7)
Ireland 0.6(0.7) 1.3(1.6) 1.5(2.1) 1.2(1.1) 63.1(60.3)
Greece 0.4(0.3) 0.9(0.8) 0.9(0.6) 0.9(1.1) 55.1(48.1)
Spain 1.3(0.7) 1.3(1.9) 0.8(1.8) 1.8(2.0) 61.2(54.8)
France 0.4(0.1) 0.4(0.6) 0.3(0.6) 0.5(0.6) 65.1(59.0)
Italy 0.2(0.1) 0.7(0.9) 0.8(0.9) 0.6(1.0) 50.6(46.8)
Luxembourg 0.3(0.3) 0.9(0.9) 1.3(1.4) 0.6(0.4) 55.4(53.5)
Netherlands 1.5(1.6) 1.2(1.4) 1.4(2.0) 0.9(0.7) 72.2(69.6)
Austria n.a.(n.a.) 0.4(0.4) 0.0(0.1) 0.8(0.7) 67.2(63.8)
Portugal 0.6(0.8) 0.8(0.6) 0.9(1.2) 0.7(0.1) 68.6(61.7)
Slovenia n.a.(n.a.) 0.5(0.5) 0.2(0.2) 0.8(0.8) 67.2(63.3)
Finland n.a.(n.a.) 0.5(0.9) 0.9(1.4) 0.1(0.4) 75.3(69.2)
Euro area 0.6(0.5) 0.7(0.9) 0.6(0.9) 0.9(0.9) 63.2(57.8)
Denmark 0.1(0.2) 0.3(0.5) 0.3(0.7) 0.2(0.3) 76.4(73.3)
Sweden n.a.(n.a.) 0.2(0.2) 0.0(0.5) 0.3(-0.1) 77.7(71.9)
United Kingdom 0.8(0.8) 0.2(0.3) 0.3(0.6) 0.1(0.0) 68.6(65.2)
Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 15 to 64 years old. The fi gures corresponding to employment rates are given in brackets. 1) 1987-1995 for Spain and Portugal. 2) 1997-2007 for Slovenia. 3) 1997-2001 for Slovenia.
93ECB
Occasional Paper No 87
June 2008
ANNEX 3
Table 28 Participation rates (and employment rates in brackets) by country and subgroup (cnt’d)
b) Subdivision according to age
15-24 years old
Average annual change (percentage points) Level (%)Trend developments Recent developments
1984-1995 1) 1996-2007 2) 1996-2001 3) 2002-2007 2007
Belgium -0.8 (-0.6) -0.1 (0.0) 0.0 (0.3) -0.1 (-0.3) 33.1 (26.8)
Germany -0.2 (0.2) -0.2 (-0.4) -0.3 (-0.3) -0.1 (-0.5) 49.7 (43.7)
Ireland -1.3 (-0.9) 0.7 (1.0) 0.8 (1.7) 0.5 (0.3) 53.1 (48.4)
Greece -0.5 (-0.5) -0.5 (-0.2) 0.0 (0.0) -0.9 (-0.3) 31.0 (24.2)
Spain -0.6 (-0.1) 0.5 (1.2) 0.1 (1.6) 0.9 (0.9) 47.8 (39.1)
France -1.5 (-1.4) 0.1 (0.3) 0.0 (0.6) 0.3 (0.0) 37.3 (29.6)
Italy -0.8 (-0.7) -0.6 (0.0) -0.4 (0.1) -0.9 (-0.1) 31.0 (25.3)
Luxembourg -1.6 (-1.5) -1.3 (-1.3) -1.1 (-1.0) -1.4 (-1.7) 26.0 (22.1)
Netherlands 1.1 (1.3) 0.9 (1.2) 1.9 (2.6) -0.1 (-0.3) 73.0 (68.6)
Austria n.a. (n.a.) -0.2 (-0.3) -1.2 (-1.1) 0.8 (0.5) 59.2 (54.5)
Portugal -2.0 (-1.4) -0.2 (-0.1) 0.6 (1.0) -0.9 (-1.3) 40.9 (34.7)
Slovenia n.a. (n.a.) -0.2 (0.2) -1.3 (-1.0) 0.7 (1.2) 40.4 (37.2)
Finland n.a. (n.a.) 1.0 (1.6) 2.2 (2.8) -0.1 (0.4) 62.1 (48.6)
Euro area -0.7 (-0.5) 0.0 (0.3) 0.0 (0.6) 0.0 (0.0) 44.0 (37.3)
Denmark 0.7 (1.1) 0.0 (0.1) -1.0 (-0.7) 0.9 (1.0) 72.6 (67.4)
Sweden n.a. (n.a.) 0.8 (0.5) 1.2 (1.6) 0.5 (-0.7) 55.1 (42.1)
United Kingdom -0.1 (0.2) -0.4 (-0.2) -0.3 (0.3) -0.4 (-0.8) 59.2 (50.8)
25-24 years oldBelgium 0.5 (0.5) 0.4 (0.5) 0.1 (0.5) 0.7 (0.5) 85.1 (79.3)
Germany 0.4 (0.3) 0.4 (0.3) 0.4 (0.4) 0.4 (0.2) 87.7 (80.8)
Ireland 0.5 (0.6) 0.8 (1.2) 1.0 (1.9) 0.6 (0.4) 82.2 (78.9)
Greece 0.5 (0.4) 0.7 (0.6) 0.6 (0.4) 0.7 (0.8) 82.0 (75.7)
Spain 1.1 (0.5) 0.7 (1.5) 0.3 (1.7) 1.1 (1.3) 82.8 (77.1)
France 0.4 (0.0) 0.1 (0.3) 0.0 (0.4) 0.3 (0.3) 87.7 (81.1)
Italy 0.1 (-0.1) 0.5 (0.7) 0.5 (0.6) 0.4 (0.8) 77.5 (73.6)
Luxembourg 0.4 (0.4) 0.8 (0.7) 1.0 (1.1) 0.5 (0.2) 82.8 (80.1)
Netherlands 0.9 (1.0) 0.7 (0.9) 0.8 (1.4) 0.6 (0.4) 87.6 (85.4)
Austria n.a. (n.a.) 0.3 (0.3) 0.3 (0.4) 0.2 (0.1) 86.7 (83.0)
Portugal 0.7 (0.7) 0.4 (0.2) 0.3 (0.7) 0.4 (-0.2) 87.7 (80.9)
Slovenia n.a. (n.a.) 0.3 (0.4) 0.2 (0.4) 0.3 (0.3) 89.9 (85.9)
Finland n.a. (n.a.) 0.2 (0.9) 0.5 (1.4) 0.0 (0.3) 88.3 (83.7)
Euro area 0.5 (0.2) 0.4 (0.6) 0.3 (0.7) 0.5 (0.5) 84.6 (79.1)
Denmark -0.2 (0.0) 0.1 (0.4) 0.1 (0.5) 0.2 (0.3) 88.5 (86.1)
Sweden n.a. (n.a.) 0.0 (0.3) -0.3 (0.3) 0.4 (0.3) 90.3 (86.3)
United Kingdom 0.4 (0.4) 0.1 (0.3) 0.0 (0.5) 0.1 (0.1) 84.5 (81.3)
Sources: EU-LFS (spring data), ECB and NBB calculations. Corresponding fi gures for employment rates between brackets.1) 1987-1995 for Spain and Portugal. 2) 1997-2007 for Slovenia. 3) 1997-2001 for Slovenia.
94ECB
Occasional Paper No 87
June 2008
Table 28 Participation rates (and employment rates in brackets) by country and subgroup (cnt’d)
b) Subdivision according to age
55-64 years old
Average annual change (percentage points) Level (%)Trend developments Recent developments
1984-1995 1) 1996-2007 2) 1996-2001 3) 2002-2007 2007
Belgium -0.5 (-0.5) 0.9 (0.9) 0.3 (0.3) 1.5 (1.4) 35.2 (33.8)
Germany 0.2 (-0.1) 1.3 (1.2) 0.0 (0.0) 2.5 (2.4) 57.9 (52.0)
Ireland -0.4 (-0.4) 1.0 (1.2) 0.8 (1.2) 1.3 (1.2) 55.5 (54.0)
Greece -0.5 (-0.5) 0.1 (0.1) -0.4 (-0.4) 0.6 (0.7) 43.6 (42.1)
Spain -0.6 (-0.5) 0.9 (1.1) 0.9 (1.2) 1.0 (0.9) 47.4 (44.8)
France -0.6 (-0.6) 0.8 (0.7) 0.2 (0.2) 1.3 (1.2) 40.5 (37.8)
Italy -0.5 (-0.5) 0.5 (0.5) -0.1 (-0.1) 1.1 (1.2) 34.8 (34.0)
Luxembourg -0.1 (-0.1) 0.9 (0.9) 0.1 (0.1) 1.6 (1.6) 34.5 (34.3)
Netherlands -0.2 (-0.1) 1.9 (1.8) 1.7 (1.7) 2.2 (2.0) 52.8 (51.0)
Austria n.a. (n.a.) 0.7 (0.7) -0.2 (-0.3) 1.6 (1.6) 38.3 (37.1)
Portugal 0.2 (0.1) 0.5 (0.4) 0.8 (0.9) 0.3 (-0.1) 54.0 (50.3)
Slovenia n.a. (n.a.) 1.4 (1.4) 0.8 (0.7) 1.9 (1.9) 36.0 (34.9)
Finland n.a. (n.a.) 1.6 (1.7) 1.7 (1.9) 1.5 (1.6) 58.8 (55.4)
Euro area -0.3 (-0.4) 0.9 (0.9) 0.2 (0.3) 1.5 (1.5) 46.4 (43.4)
Denmark 0.0 (-0.1) 0.6 (0.8) 0.9 (1.2) 0.4 (0.4) 61.3 (58.7)
Sweden n.a. (n.a.) 0.5 (0.7) 0.4 (0.7) 0.6 (0.6) 72.9 (69.9)
United Kingdom -0.1 (0.0) 0.7 (0.8) 0.4 (0.8) 0.9 (0.9) 59.3 (57.4)
Sources: EU-LFS (spring data), ECB and NBB calculations. Corresponding fi gures for employment rates between brackets.1) 1987-1995 for Spain and Portugal. 2) 1997-2007 for Slovenia. 3) 1997-2001 for Slovenia.
c) Subdivision according to education level
Average annual change (percentage points) Level (%)Trend developments Recent developments
1996-2007 1) 1996-2001 2) 2002-2007 2007
LowBelgium 0.2 (0.2) 0.0 (0.3) 0.4 (0.2) 55.9 (49.3)
Germany 0.9 (0.6) 0.7 (0.7) 1.1 (0.4) 66.4 (54.7)
Ireland 0.4 (0.8) 0.3 (1.3) 0.4 (0.3) 62.8 (58.7)
Greece 0.4 (0.3) 0.3 (0.1) 0.5 (0.6) 64.5 (59.9)
Spain 0.6 (1.2) 0.4 (1.3) 0.9 (1.0) 66.5 (60.9)
France 0.1 (0.3) 0.2 (0.4) 0.1 (0.1) 64.6 (57.4)
Italy 0.2 (0.4) 0.2 (0.2) 0.3 (0.5) 56.1 (52.7)
Luxembourg 0.4 (0.3) 0.5 (0.6) 0.2 (0.0) 60.7 (58.1)
Netherlands 0.6 (0.8) 0.8 (1.5) 0.5 (0.3) 64.1 (61.4)
Austria 0.0 (-0.1) -0.8 (-0.9) 0.9 (0.7) 61.6 (56.4)
Portugal 0.5 (0.3) 0.7 (1.0) 0.3 (-0.3) 77.5 (71.2)
Slovenia 0.5 (0.5) 1.1 (1.0) 0.0 (0.1) 61.5 (57.2)
Finland 0.0 (0.5) 0.2 (0.8) -0.1 (0.2) 65.1 (59.3)
Euro area 0.4 (0.6) 0.3 (0.7) 0.5 (0.5) 63.5 (57.6)
Denmark 0.5 (0.9) 0.3 (0.8) 0.8 (0.9) 70.3 (67.5)
Sweden -1.3 (-0.9) -2.3 (-1.5) -0.2 (-0.4) 71.5 (66.6)
United Kingdom -0.4 (-0.1) -0.6 (-0.1) -0.1 (-0.1) 68.6 (64.4)
Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 25 to 64 years old. EU-LFS data concerning education level are only available from 1992 onwards. The fi gures corresponding to employment rates are given in brackets. 1) 1997-2007 for the Netherlands and Slovenia. 2) 1997-2001 for the Netherlands and Slovenia.
95ECB
Occasional Paper No 87
June 2008
ANNEX 3
Table 28 Participation rates (and employment rates in brackets) by country and subgroup (cnt’d)
c) Subdivision according to education level
Average annual change (percentage points) Level (%)Trend developments Recent developments
1996-2007 1) 1996-2001 2) 2002-2007 2007
MediumBelgium 0.0 (0.1) 0.0 (0.4) 0.1 (-0.2) 78.8 (73.9)
Germany 0.5 (0.5) 0.2 (0.2) 0.8 (0.8) 81.6 (74.8)
Ireland 0.7 (0.9) 1.2 (1.7) 0.2 (0.0) 80.2 (77.5)
Greece 0.5 (0.5) 0.7 (0.5) 0.3 (0.6) 75.7 (69.6)
Spain 0.2 (1.0) -0.3 (1.2) 0.7 (0.8) 81.9 (76.6)
France -0.2 (0.0) -0.2 (0.1) -0.2 (-0.2) 80.3 (74.9)
Italy 0.1 (0.3) 0.1 (0.2) 0.1 (0.5) 78.1 (75.1)
Luxembourg 0.4 (0.4) 0.7 (0.8) 0.1 (0.0) 75.9 (74.1)
Netherlands 0.4 (0.6) 0.7 (1.2) 0.2 (0.1) 82.6 (80.4)
Austria 0.0 (0.0) -0.2 (-0.1) 0.3 (0.2) 78.5 (75.7)
Portugal 0.2 (0.2) 0.4 (0.9) 0.0 (-0.5) 85.4 (79.7)
Slovenia 0.0 (0.0) -0.3 (-0.1) 0.1 (0.1) 78.6 (75.1)
Finland 0.1 (0.7) 0.4 (1.3) -0.2 (0.2) 81.8 (77.0)
Euro area 0.2 (0.4) 0.1 (0.3) 0.4 (0.4) 80.4 (75.3)
Denmark 0.1 (0.4) 0.1 (0.5) 0.2 (0.3) 84.2 (82.2)
Sweden -0.3 (0.0) -0.9 (-0.2) 0.2 (0.2) 87.1 (83.3)
United Kingdom 0.0 (0.2) 0.1 (0.7) -0.1 (-0.2) 84.2 (81.1)
HighBelgium 0.1 (0.1) 0.0 (0.1) 0.2 (0.1) 87.8 (84.9)
Germany 0.2 (0.3) -0.1 (0.0) 0.5 (0.5) 89.8 (86.5)
Ireland 0.2 (0.3) 0.3 (0.7) 0.1 (0.0) 89.0 (87.0)
Greece 0.3 (0.3) 0.2 (0.0) 0.3 (0.5) 88.4 (83.3)
Spain 0.1 (0.8) -0.2 (1.0) 0.4 (0.6) 88.8 (84.6)
France -0.1 (0.1) 0.0 (0.3) -0.1 (-0.1) 87.2 (83.0)
Italy -0.3 (-0.1) -0.2 (0.1) -0.4 (-0.2) 84.1 (80.6)
Luxembourg 0.3 (0.1) 0.5 (0.5) 0.0 (-0.3) 86.5 (84.0)
Netherlands 0.3 (0.4) 0.3 (0.8) 0.2 (0.2) 89.6 (88.1)
Austria -0.2 (-0.2) -0.4 (-0.3) 0.1 (0.0) 88.9 (86.7)
Portugal 0.0 (-0.2) 0.0 (0.2) 0.0 (-0.6) 92.5 (86.9)
Slovenia 0.3 (0.3) -0.1 (0.0) 0.6 (0.5) 91.6 (89.1)
Finland 0.0 (0.3) 0.2 (0.7) -0.2 (-0.1) 88.2 (85.2)
Euro area 0.0 (0.2) -0.1 (0.3) 0.2 (0.2) 88.3 (84.7)
Denmark -0.1 (0.1) -0.2 (0.1) 0.0 (0.1) 90.2 (87.6)
Sweden -0.1 (-0.1) -0.7 (-0.4) 0.4 (0.2) 91.9 (88.6)
United Kingdom 0.0 (0.1) 0.0 (0.3) 0.0 (0.0) 89.8 (88.0)
Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 25 to 64 years old. EU-LFS data concerning education level are only available from 1992 onwards. The fi gures corresponding to employment rates are given in brackets. 1) 1997-2007 for the Netherlands and Slovenia. 2) 1997-2001 for the Netherlands and Slovenia.
96ECB
Occasional Paper No 87
June 2008
Table 28 Participation rates (and employment rates in brackets) by country and subgroup (cnt’d)
d) Subdivision according to nationality
Nationals
Average annual change (percentage points) Level (%)Trend
developments Recent developments1996-2007 1996-2001 2002-2007 1) 2007
Belgium 0.4 (0.4) 0.3 (0.5) 0.4 (0.3) 67.0 (62.4)
Germany 0.5 (0.4) 0.2 (0.2) 0.8 (0.7) 76.7 (70.6)
Ireland n.a. (n.a.) n.a. (n.a.) 0.6 (0.5) 71.4 (68.2)
Greece 0.6 (0.6) 0.5 (0.3) 0.6 (0.8) 66.6 (61.1)
Spain 0.8 (1.5) 0.6 (1.8) 1.1 (1.3) 70.5 (65.3)
France 0.2 (0.3) 0.2 (0.5) 0.2 (0.1) 70.0 (64.2)
Italy n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) 61.9 (58.4)
Luxembourg 0.4 (0.3) 0.5 (0.6) 0.3 (0.0) 61.7 (59.4)
Netherlands n.a. (n.a.) n.a. (n.a.) 0.4 (0.3) 79.1 (76.7)
Austria 0.3 (0.3) -0.1 (-0.1) 0.7 (0.6) 74.2 (71.3)
Portugal n.a. (n.a.) n.a. (n.a.) 0.3 (-0.2) 73.4 (67.5)
Slovenia n.a. (n.a.) n.a. (n.a.) 0.6 (0.8) 71.7 (68.4)
Finland 0.4 (1.0) 0.8 (1.6) 0.0 (0.4) 77.4 (71.6)
Euro area 0.3 (0.5) 0.4 (0.8) 0.2 (0.3) 70.9 (65.9)
Denmark 0.1 (0.3) 0.0 (0.4) 0.2 (0.3) 81.2 (78.5)
Sweden n.a. (n.a.) n.a. (n.a.) 0.3 (0.0) 80.4 (75.1)
United Kingdom 0.0 (0.2) 0.0 (0.5) 0.0 (-0.1) 75.2 (71.4)
Other EU15 citizensBelgium 0.5 (0.9) 0.2 (1.0) 0.8 (0.7) 69.1 (62.4)
Germany 0.2 (0.2) 0.1 (0.3) 0.3 (0.0) 76.7 (69.6)
Ireland n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.)
Greece 0.6 (-0.6) 0.4 (-0.6) -1.6 (-0.6) 53.6 (48.8)
Spain 0.7 (1.1) 0.5 (1.8) 1.0 (0.4) 70.2 (62.7)
France 0.1 (0.3) -0.1 (0.3) 0.4 (0.4) 72.8 (67.8)
Italy n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) 60.0 (57.9)
Luxembourg 0.4 (0.4) 0.7 (0.9) 0.1 (-0.1) 71.6 (69.3)
Netherlands n.a. (n.a.) n.a. (n.a.) 0.2 (0.2) 78.0 (75.8)
Austria 0.8 (0.5) 0.4 (0.0) 1.1 (1.0) 77.2 (73.1)
Portugal n.a. (n.a.) n.a. (n.a.) 1.7 (1.3) 74.6 (69.3)
Slovenia n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.)
Finland 1.5 (2.2) 2.7 (4.9) 0.3 (-0.4) 85.1 (77.8)
Euro area 0.2 (0.3) 0.0 (0.5) 0.4 (0.2) 73.7 (67.6)
Denmark 0.1 (0.3) 1.4 (1.1) -1.2 (-0.5) 76.6 (73.8)
Sweden n.a. (n.a.) n.a. (n.a.) 0.6 (0.3) 75.4 (70.8)
United Kingdom 0.4 (0.6) 0.1 (1.0) 0.7 (0.3) 76.9 (71.5)
Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 15 to 64 years old. EU-LFS data concerning nationality are only available from 1995 onwards. The fi gures corresponding to employment rates are given in brackets. 1) 2003-07 for Slovenia.
97ECB
Occasional Paper No 87
June 2008
ANNEX 3
Table 28 Participation rates (and employment rates in brackets) by country and subgroup (cnt’d)
d) Subdivision according to nationality
Non EU15 citizens
Average annual change (percentage points) Level (%)Trend
developments Recent developments1996-2007 1996-2001 2002-2007 1) 2007
Belgium 1.1 (1.1) 0.0 (0.8) 2.1 (1.5) 55.7 (40.6)
Germany 0.0 (-0.1) -0.3 (0.0) 0.3 (-0.2) 63.8 (51.5)
Ireland n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) n.a. (n.a.)
Greece 0.1 (0.5) 0.3 (0.7) 0.0 (0.4) 73.8 (67.9)
Spain 0.7 (1.4) 1.2 (2.2) 0.2 (0.5) 79.7 (70.0)
France 0.1 (0.4) 0.2 (0.6) 0.1 (0.2) 59.3 (44.9)
Italy n.a. (n.a.) n.a. (n.a.) n.a. (n.a.) 73.0 (67.4)
Luxembourg 0.9 (0.5) 1.5 (1.2) 0.2 (-0.2) 64.9 (57.7)
Netherlands n.a. (n.a.) n.a. (n.a.) 0.9 (0.4) 57.6 (51.8)
Austria -0.7 (-1.0) -0.3 (-0.5) -1.2 (-1.5) 67.8 (59.5)
Portugal n.a. (n.a.) n.a. (n.a.) 1.3 (0.4) 82.8 (70.6)
Slovenia n.a. (n.a.) n.a. (n.a.) 2.1 (1.3) 65.7 (59.7)
Finland 0.7 (0.9) 1.6 (1.0) -0.2 (0.9) 67.8 (53.6)
Euro area 0.6 (0.8) 0.2 (0.5) 1.1 (1.1) 69.6 (59.3)
Denmark 0.2 (0.9) -0.9 (0.7) 1.4 (1.2) 61.1 (53.9) Sweden n.a. (n.a.) n.a. (n.a.) 0.3 (-0.2) 65.1 (52.3)
United Kingdom 0.8 (1.2) -0.1 (0.6) 1.7 (1.9) 71.0 (65.2)
Non EU15 citizens: level (%) in 2007
Citizens of the 12 new EU member states
Non EU27 citizens
Belgium 71.1 (62.1) 53.2 (37.1)
Germany 73.3 (63.7) 62.4 (49.7)
Ireland n.a. (n.a.) n.a. (n.a.)
Greece 71.7 (65.6) 74.2 (68.4)
Spain 82.4 (72.9) 79.0 (69.2)
France 68.0 (53.6) 58.9 (44.5)
Italy 76.7 (71.1) 72.2 (66.6)
Luxembourg 63.4 (61.4) 65.4 (56.4)
Netherlands 72.3 (67.7) 56.3 (50.4)
Austria 75.1 (69.6) 66.3 (57.3)
Portugal 72.9 (67.8) 83.5 (70.8)
Slovenia n.a. (n.a.) 65.1 (59.0)
Finland 80.0 (74.4) 63.9 (47.0)
Euro area 77.1 (68.7) 68.3 (57.7)
Denmark 76.3 (71.8) 60.3 (52.9)
Sweden 74.1 (56.0) 63.9 (51.9)
United Kingdom 84.7 (79.6) 66.7 (60.7)
Sources: EU-LFS (spring data), ECB and NBB calculations. Notes: 15 to 64 years old. EU-LFS data concerning nationality are only available from 1995 onwards. The fi gures corresponding to employment rates are given in brackets. 1) 2003-07 for Slovenia.
98ECB
Occasional Paper No 87
June 2008
Table 29 Developments in the percentage of nationals and non-nationals in the working age population, participation and employment, by country
% of working age population
Country Nationals Other EU15 citizens Non EU15 citizens
average annual change level average annual change level average annual change level1996-2001 2002-2007 2007 1996-2001 2002-2007 2007 1996-2001 2002-2007 2007
Belgium -0.1 0.1 90.9 0.1 -0.1 5.3 0.0 0.1 3.8
Germany -0.1 -0.1 89.6 0.0 0.0 2.8 0.1 0.1 7.6
Ireland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Spain -0.5 -1.5 87.1 0.1 0.1 1.5 0.4 1.4 11.4
Greece -0.4 -0.4 93.8 0.0 0.0 0.3 0.4 0.4 6.0
France 0.0 0.2 94.1 0.0 0.0 2.1 0.0 -0.1 3.8
Italy n.a. n.a. 94.2 n.a. n.a. 0.3 n.a. n.a. 5.5
Luxembourg -0.8 -0.3 57.9 0.5 0.3 37.7 0.3 0.0 4.4
Netherlands n.a. 0.1 95.7 n.a. 0.0 1.6 n.a. -0.1 2.7
Austria -0.1 -0.3 88.7 0.1 0.1 2.2 0.0 0.2 9.1
Portugal n.a. -0.3 96.3 n.a. 0.0 0.4 n.a. 0.3 3.3
Slovenia n.a. 0.0 99.3 n.a. n.a. n.a. n.a. 0.0 0.7Finland -0.1 -0.1 98.0 0.0 0.0 0.4 0.1 0.1 1.6
Euro area -0.1 -0.2 91.9 0.0 -0.1 1.8 0.1 0.3 6.3
Denmark -0.2 -0.3 94.6 0.0 0.0 1.1 0.2 0.3 4.4
Sweden n.a. 0.1 95.0 n.a. 0.0 2.1 n.a. 0.0 2.9
United Kindgom -0.2 -0.4 92.2 0.0 0.0 1.8 0.2 0.4 6.0
Participation rates Belgium 0.3 0.4 67.0 0.2 0.8 69.1 0.0 2.1 55.7
Germany 0.2 0.8 76.7 0.1 0.3 76.7 -0.3 0.3 63.8
Ireland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Spain 0.6 1.1 70.5 0.5 1.0 70.2 1.2 0.2 79.7
Greece 0.5 0.6 66.6 0.4 -1.6 53.6 0.3 0.0 73.8
France 0.2 0.2 70.0 -0.1 0.4 72.8 0.2 0.1 59.3
Italy n.a. n.a. 61.9 n.a. n.a. 60.0 n.a. n.a. 73.0
Luxembourg 0.5 0.3 61.7 0.7 0.1 71.6 1.5 0.2 64.9
Netherlands n.a. 0.4 79.1 n.a. 0.2 78.0 n.a. 0.9 57.6
Austria -0.1 0.7 74.2 0.4 1.1 77.2 -0.3 -1.2 67.8
Portugal n.a. 0.3 73.4 n.a. 1.7 74.6 n.a. 1.3 82.8
Slovenia n.a. 0.5 71.7 n.a. n.a. n.a n.a. 1.8 65.7Finland 0.8 0.0 77.4 2.7 0.3 85.1 1.6 -0.2 67.8
Euro area 0.4 0.2 70.9 0.0 0.4 73.7 0.2 1.1 69.6
Denmark 0.0 0.2 81.2 1.4 -1.2 76.6 -0.9 1.4 61.1
Sweden n.a. 0.3 80.4 n.a. 0.6 75.4 n.a. 0.3 65.1
United Kindgom 0.0 0.0 75.2 0.1 0.7 76.9 -0.1 1.7 71.0
Source: EU-LFS (spring data).
Notes: IE: data available for 1998-2004 only. IT: for 2005-07 only. NL and PT: data start 1999 SI: data start 2002 SE: data start 1997.
The non-national population is separated into non-national EU15 citizens and non-national non-EU15 citizens. Numbers in italics are
based on fi gures smaller than the Eurostat reliability limit.
Chart 12 Annual and weekly hours worked in the EU15 countries and the United States in 2005
30 32 34 36 38 40 42 44
2,100
2,000
1,900
1,800
1,700
1,600
1,500
1,400
1,300
2,100
2,000
1,900
1,800
1,700
1,600
1,500
1,400
1,300
NL
DK
DE
AT
UK
SE FR
LU
BEEA
FI
IT
IE
PTES
USGR
x-axis: usual weekly hours worked
y-axis: annual hours worked
Sources: EU-LFS (spring data) and OECD.
99ECB
Occasional Paper No 87
June 2008
ANNEX 3
Table 29 Developments in the percentage of nationals and non-nationals in the working age population, participation and employment, by country (cnt’d)
% of working age population
Country Nationals Other EU15 citizens Non EU15 citizens
average annual change level average annual change level average annual change level1996-2001 2002-2007 2007 1996-2001 2002-2007 2007 1996-2001 2002-2007 2007
Employment rates 2007
Belgium 0.5 0.3 62.4 1.0 0.7 62.4 0.8 1.5 40.6
Germany 0.2 0.7 70.6 0.3 0.0 69.6 0.0 -0.2 51.5
Ireland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Spain 1.8 1.3 65.3 1.8 0.4 62.7 2.2 0.5 70.0
Greece 0.3 0.8 61.1 -0.6 -0.6 48.8 0.7 0.4 67.9
France 0.5 0.1 64.2 0.3 0.4 67.8 0.6 0.2 44.9
Italy n.a. n.a. 58.4 n.a. n.a. 57.9 n.a. n.a. 67.4
Luxembourg 0.6 0.0 59.4 0.9 -0.1 69.3 1.2 -0.2 57.7
Netherlands n.a. 0.3 76.7 n.a. 0.2 75.8 n.a. 0.4 51.8
Austria -0.1 0.6 71.3 0.0 1.0 73.1 -0.5 -1.5 59.5
Portugal n.a. -0.2 67.5 n.a. 1.3 69.3 n.a. 0.4 70.6
Slovenia n.a. 0.7 68.4 n.a. n.a. n.a. n.a. 1.0 59.7Finland 1.6 0.4 71.6 4.9 -0.4 77.8 1.0 0.9 53.6
Euro area 0.8 0.3 65.9 0.5 0.2 67.6 0.5 1.1 59.3
Denmark 0.4 0.3 78.5 1.1 -0.5 73.8 0.7 1.2 53.9
Sweden n.a. 0.0 75.1 n.a. 0.3 70.8 n.a. -0.2 52.3
United Kindgom 0.5 -0.1 71.4 1.0 0.3 71.5 0.6 1.9 65.2
Source: EU-LFS (spring data).
Notes: IE: data available for 1998-2004 only. IT: for 2005-07 only. NL and PT: data start 1999 SI: data start 2002 SE: data start 1997.
The non-national population is separated into non-national EU15 citizens and non-national non-EU15 citizens. Numbers in italics are
based on fi gures smaller than the Eurostat reliability limit.
Table 30 Educational Attainment of 25-54 year olds
(% of total population)
1992 1999 2007 Country Low Medium High Low Medium High Low Medium High
Belgium 44.5 31.7 23.7 37.6 33.2 29.2 27.6 38.4 34.0
Germany 16.5 61.1 22.4 17.4 58.6 24.0 14.2 61.0 24.8
Ireland 54.1 27.7 18.2 40.6 37.5 21.9 27.2 37.5 35.3
Greece 56.8 28.5 14.7 42.7 37.9 19.3 34.2 41.5 24.2
Spain 71.6 13.5 14.9 58.5 17.5 24.1 44.5 23.7 31.8
France n.a. n.a. n.a. 34.6 42.4 22.9 27.2 43.1 29.6
Italy 62.3 30.1 7.6 51.5 38.1 10.4 42.8 42.5 14.7
Luxembourg 62.4 24.3 13.3 34.5 45.8 19.7 30.6 39.3 30.1
Netherlands n.a. n.a. n.a. 32.2 44.0 23.8 23.3 44.5 32.1
Austria n.a. n.a. n.a. 21.4 63.6 15.0 17.8 64.0 18.2
Portugal 76.8 11.2 12.0 77.9 12.1 10.0 68.7 16.1 15.3
Slovenia n.a. n.a. n.a. 22.3 61.3 16.3 14.9 60.3 24.8
Finland n.a. n.a. n.a. 22.6 43.7 33.8 14.6 47.2 38.2
Euro area 45.2 38.9 15.9 37.6 41.8 20.6 30.5 44.2 25.4
Denmark 21.4 57.8 20.8 18.0 53.7 28.3 21.9 43.9 34.2
Sweden n.a. n.a. n.a. 19.3 50.2 30.5 11.9 55.0 33.1
United Kingdom 48.8 31.4 19.8 35.3 36.4 28.3 25.6 41.2 33.2
Source: EU – LFS (spring data).
100ECB
Occasional Paper No 87
June 2008
Table 31 Educational Attainment of Population (% of cohort population by highest level of education attained)
LOW EDUCATION: This Table shows the percentage of a cohort population with a low level of education,
e.g. Belgium 20-24 = 25.6 => that 25.8 % of the 20-24 year old population in Belgium has a low level of education
1992 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 25.6 33.6 38.9 41.6 46.3 53.4 61.0 68.3 74.4
Germany 17.6 12.7 13.8 14.1 16.2 18.6 23.9 30.7 35.7
Ireland 32.5 41.8 46.3 51.9 60.2 63.3 68.2 73.2 77.6
Greece 28.8 37.4 45.1 52.4 62.2 68.1 75.4 80.2 84.1
Spain 47.3 54.3 63.1 71.4 79.3 85.0 87.5 90.9 91.9
France n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Italy 44.9 52.1 54.4 57.0 63.9 72.2 78.4 84.3 86.4
Luxembourg 54.4 55.2 59.7 61.9 64.0 68.2 69.5 76.9 77.3
Netherlands n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Austria n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Portugal 65.0 65.3 70.2 75.1 80.6 84.2 88.0 91.2 92.1
Slovenia n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Finland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Euro area 34.8 35.9 39.4 42.7 48.3 53.7 54.5 63.5 68.3
Denmark 21.3 12.8 15.2 19.3 24.1 24.7 35.6 44.3 48.3
Sweden n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
United Kingdom 42.9 46.4 46.8 46.2 47.7 51.4 56.7 61.3 58.5
1999 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 23.8 22.5 28.8 35.1 41.9 45.9 53.3 59.6 68.8
Germany 25.4 16.7 15.4 16.5 16.8 18.4 21.7 25.0 32.3
Ireland 18.0 24.9 31.0 37.5 43.4 53.2 60.0 63.0 70.8
Greece 21.4 26.5 32.9 39.0 46.3 53.5 63.3 73.2 79.0
Spain 34.8 42.7 49.1 55.3 62.7 71.4 79.0 84.0 88.9
France 20.0 22.3 26.4 33.0 38.4 42.3 46.7 54.8 63.9
Italy 33.7 40.4 47.3 48.9 50.9 58.4 67.3 74.6 81.1
Luxembourg 28.8 32.3 31.1 32.7 33.9 39.8 40.6 48.7 57.4
Netherlands 27.7 24.4 26.7 29.7 32.9 38.8 42.6 48.0 51.5
Austria 15.3 15.0 16.7 18.7 23.2 29.0 29.5 35.6 47.8
Portugal 59.9 64.6 74.7 78.7 80.0 85.0 87.9 91.3 93.5
Slovenia 14.2 12.4 18.2 18.9 26.5 27.7 31.6 36.9 46.1
Finland 13.2 15.5 13.5 15.5 21.0 27.7 38.9 48.7 59.1
Euro area 28.4 29.3 32.0 35.1 38.7 43.4 50.2 53.3 60.7
Denmark 26.8 10.6 14.8 20.2 20.5 20.2 22.0 25.2 36.4
Sweden 13.7 12.7 12.5 17.1 21.2 23.1 28.6 34.4 41.3
United Kingdom 24.7 30.7 34.8 35.5 34.6 35.7 41.0 48.2 43.8
2007 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 17.8 17.6 20.4 23.1 27.8 33.7 41.5 45.0 53.8
Germany 28.1 14.8 15.0 13.6 13.5 13.8 15.2 18.8 20.2
Ireland 13.5 14.3 20.3 24.7 31.6 35.8 45.1 55.9 60.9
Greece 17.6 23.4 25.9 29.4 35.6 42.5 50.9 59.1 68.0
Spain 39.4 34.6 36.2 41.5 46.1 53.8 60.5 68.3 75.9
France 18.4 15.7 18.8 24.2 28.8 34.5 40.6 45.1 49.5
Italy 24.0 28.0 35.3 42.0 46.9 50.4 53.8 61.8 72.1
Luxembourg 29.7 18.0 24.1 29.3 35.8 35.4 38.8 46.9 48.8
Netherlands 24.4 16.4 17.8 21.4 24.3 27.2 31.2 36.9 42.0
Austria 16.1 13.8 13.5 16.0 17.5 21.1 25.0 29.0 29.5
Portugal 47.1 51.2 60.1 69.1 75.7 77.2 81.4 85.1 88.8
Slovenia 9.1 4.5 10.0 14.4 15.7 18.7 26.2 27.5 29.2
Finland 18.2 10.5 9.8 14.8 12.4 16.0 22.5 30.7 40.3
Euro area 26.1 22.5 26.1 28.8 31.1 34.9 39.1 44.8 51.0
Denmark 28.7 15.0 14.7 18.4 23.2 28.2 30.9 31.5 39.2
Sweden 13.0 10.3 8.0 9.0 11.1 14.7 18.4 22.4 29.4
United Kingdom 21.7 19.6 20.4 26.6 27.5 29.0 29.3 34.1 31.1
101ECB
Occasional Paper No 87
June 2008
ANNEX 3
Table 31 Educational Attainment of Population (% of cohort population by highest level of education attained) (cnt’d)
HIGH EDUCATION: This Table shows the percentage of a cohort population with a high level of education.
1992 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 16.2 29.0 26.6 24.6 22.7 19.7 16.2 11.0 8.2
Germany 5.1 16.3 24.1 26.0 24.9 24.2 20.2 17.0 14.6
Ireland 18.1 22.3 20.0 18.8 17.0 15.7 13.3 10.7 8.9
Greece 7.4 20.8 17.7 16.8 13.6 10.9 8.3 6.0 4.7
Spain 11.8 22.3 18.6 15.0 11.8 9.1 7.6 5.7 4.9
France n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Italy 1.1 5.8 8.6 10.2 9.2 6.9 5.2 4.0 4.0
Luxembourg 9.1 14.4 12.9 13.6 14.6 13.2 10.4 6.7 5.8
Netherlands n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Austria n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Portugal 3.9 13.7 15.1 15.0 11.3 8.7 7.4 5.5 4.8
Slovenia n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Finland n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Euro area 6.0 15.0 17.9 18.3 16.4 14.5 12.8 9.8 8.2
Denmark 0.6 11.6 23.1 25.8 24.7 21.3 19.3 12.6 10.6
Sweden n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
United Kingdom 12.7 20.8 21.3 21.4 20.5 18.2 15.5 14.1 13.5
1999 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 15.4 37.2 32.0 30.5 27.2 25.3 22.4 18.4 13.0
Germany 4.2 17.6 24.5 25.2 26.0 25.8 23.8 21.7 17.6
Ireland 20.9 30.8 24.7 21.3 20.3 17.0 14.8 14.0 11.0
Greece 6.0 21.6 25.4 20.8 18.1 16.2 12.0 9.2 6.5
Spain 22.6 35.2 29.3 25.4 19.5 16.6 12.7 10.0 6.8
France 25.2 33.6 26.2 21.7 20.0 18.7 16.9 13.4 9.7
Italy 1.4 9.1 10.8 11.0 11.9 10.8 8.8 7.0 4.8
Luxembourg 15.9 20.6 21.7 16.9 18.4 21.2 19.3 15.6 8.1
Netherlands 7.9 24.9 25.4 23.7 25.9 23.2 19.4 18.3 15.4
Austria 5.2 14.8 16.3 16.4 14.9 13.8 12.9 12.4 9.3
Portugal 5.2 13.7 11.1 9.2 9.9 8.4 6.6 5.1 3.8
Slovenia 3.3 19.5 15.7 19.4 14.2 14.3 14.9 11.7 11.1
Finland 9.3 35.4 39.2 35.7 35.1 29.9 28.6 24.0 17.1
Euro area 11.7 23.0 22.6 21.2 20.4 19.0 16.4 14.6 11.2
Denmark 3.5 26.1 31.1 28.8 29.0 27.4 27.2 21.3 16.2
Sweden 20.1 31.9 31.8 31.1 29.1 31.8 27.7 22.7 18.3
United Kingdom 22.0 30.6 27.9 27.8 29.1 29.3 24.9 21.0 21.4
2007 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64Belgium 18.1 38.4 39.3 37.7 34.0 29.9 25.8 23.9 20.1
Germany 3.3 18.8 26.2 25.6 25.7 25.4 26.0 23.2 22.1
Ireland 25.3 45.1 41.9 37.8 30.5 26.9 23.1 18.4 16.2
Greece 9.6 28.0 26.8 25.3 25.1 20.7 18.8 15.2 10.9
Spain 20.1 38.3 39.6 34.4 29.9 24.3 20.1 17.4 14.0
France 26.2 42.8 40.7 31.4 24.6 20.9 18.7 17.4 15.6
Italy 7.4 19.1 18.5 16.0 12.1 10.8 11.6 11.1 7.4
Luxembourg 10.4 40.1 36.3 32.3 23.1 22.7 28.5 19.0 20.2
Netherlands 14.5 36.7 36.4 31.7 29.4 30.4 29.6 27.5 23.4
Austria 3.1 16.4 21.6 18.8 17.8 17.3 17.0 13.9 12.9
Portugal 7.1 22.0 19.7 14.6 12.5 11.1 10.5 7.8 6.3
Slovenia 2.4 32.1 31.6 24.4 20.5 22.3 17.9 15.7 15.8
Finland 1.0 26.3 46.3 42.8 41.9 38.5 34.0 28.5 27.0
Euro area 12.8 29.5 30.8 26.6 23.6 21.5 20.5 18.3 15.7
Denmark 4.6 36.9 41.9 35.3 32.0 30.1 29.3 24.4 21.2
Sweden 10.1 39.0 40.6 32.9 28.9 29.1 28.9 27.8 24.5
United Kingdom 21.0 36.4 38.1 32.8 31.9 30.5 30.3 26.4 23.8
102ECB
Occasional Paper No 87
June 2008
Table 32 Educational attainment expressed in average number of years in formal education 2004
25-to-64-year-old populationMales Females
Total Males Females 25-34 35-44 45-54 54-64 25-34 35-44 45-54 54-64
Belgium 11.3 11.4 11.4 12.4 11.7 11.1 10.3 12.8 11.9 10.7 9.5
Germany 13.4 13.7 13.2 13.6 13.8 13.8 13.7 13.5 13.4 13.2 12.5
Ireland 13.0 12.9 13.1 14.0 13.4 12.3 11.2 14.5 13.6 12.5 11.4
Greece 10.9 11.0 10.7 11.9 11.7 10.9 9.4 12.6 11.7 10.0 8.2
Spain 10.6 10.6 10.6 11.9 11.2 10.1 8.9 12.5 11.4 9.7 8.0
France 11.6 11.7 11.4 12.8 12.1 11.3 10.3 13.1 12.0 10.7 9.6
Italy 10.1 10.2 10.0 11.2 10.5 10.0 8.7 11.7 10.7 9.5 7.6
Luxembourg 13.3 13.6 13.0 14.2 13.5 13.5 13.1 14.1 13.3 12.6 11.6
Netherlands 11.2 11.4 11.1 12.0 11.5 11.3 10.6 12.5 11.4 10.5 9.8
Austria 12.0 12.3 11.7 12.4 12.4 12.2 12.0 12.3 12.0 11.4 10.8
Portugal 8.5 8.3 8.7 9.3 8.4 7.8 7.3 10.3 8.8 7.9 7.2
Slovenia n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Finland 11.2 10.9 11.4 12.5 12.3 10.5 8.5 13.5 13.0 11.2 8.5
Euro area 11.4 11.5 11.4 12.4 11.9 11.2 10.3 12.8 11.9 10.8 9.6
Denmark 13.4 13.5 13.3 13.6 13.6 13.4 13.6 13.6 13.3 13.3 13.0
Sweden 12.6 12.4 12.8 13.1 12.7 12.2 11.3 13.6 13.0 12.7 11.8
United Kingdom 12.6 12.7 12.4 13.0 12.6 12.7 12.4 12.9 12.4 12.3 12.0
Source: OECD “Education at a Glance, 2006” (Table A1.5). Note: Euro area in this table is calculated as a simple average of the available Euro area countries.
Chart 13 Correlation between annual expenditure per student and aggregate Pisa scores
GRPT
ESFR
IT
LU
ATBEIEDE
NL
FI
1,350
1,400
1,450
1,500
1,550
1,600
1,650
1,700
1,350
1,400
1,450
1,500
1,550
1,600
1,650
1,700
12,0002,000 4,000 6,000 8,000 10,000
y-axis: aggregate 2006 Pisa scc
x-axis: annual expenditure per primary student
IEDE
BE
AT
FRESPT
ITGR
LU
NL
FI
1,350
1,400
1,450
1,500
1,550
1,600
1,650
1,700
1,350
1,400
1,450
1,500
1,550
1,600
1,650
1,700
16,0002,000 4,000 6,000 8,000 10,000 12,000 14,000
y-axis: aggregate 2006 Pisa score
x-axis: annual expenditure per secondary student
Source: OECD (2007d) and ECB calculations.
103ECB
Occasional Paper No 87
June 2008
ANNEX 3
Table 33 Private internal rates of return for an individual obtaining a university-level degree, ISCED 5/6 (2003)
Rate of return when the individual immediately aquires the next higher
level of education
Rate of return when the individual, at age 40, begins the next higher level of education in full time studies, and the
individual bears 1):
Direct costs of foregone earningsNo direct costs but foregone earnings
Males % Females % Males % Females % Males % Females %
Belgium 10.7 15.2 20.0 28.2 21.1 32.2
Denmark 8.3 8.1 12.4 10.2 12.5 10.5
Finland 16.7 16.0 16.2 13.2 16.4 13.4
Sweden 8.9 8.2 10.4 8.2 10.8 8.7
United Kingdom 16.8 19.6 11.4 14.9 12.5 16.8
United States 14.3 13.1 12.9 9.7 15.1 13.0
Source: OECD, “Education at a Glance, 2007”. 1) Private internal rate of return: additions to after-tax earnings that result from higher education net of the additional private costs of education attainment (private expenditures and foregone earnings). Living expenses are excluded from these private expenses. Direct costs are costs of tuition as reported by the national authorities. Foregone earnings are net of taxes.
Table 34 Unemployment rates and mismatch by type of education (tertiary education only) in the euro area and the euro area countries 1)
Unemployment rate in 2006 (%)mismatch
(range)
Country General
Teacher training
and educa-
tion
Humani-ties
langu ages and
art
Social science
business, law
Science math
compu-ting
Engin eering
manufac-turing
Agri-culture
Health and
welfareSer-
vices
Weighted average
of sub groups 2) 2006
Change (p.p.) 5)
2004-2006 3)
Belgium 13.1 2.3 6.2 4.4 4.2 3.5 3.1 2.2 6.1 3.8 4.0 -3.1 -1.6Germany n.a. 3.9 4.3 4.2 5.0 4.9 3.5 3.4 3.6 4.4 1.6 -1.9 -0.6Ireland 4) 2.1 1.6 4.0 1.6 2.6 2.0 0.3 1.2 2.1 2.0 3.7 n.a. n.a Greece n.a. 3.9 8.1 7.6 5.2 4.8 7.5 7.0 4.0 6.3 4.2 -1.0 -0.3Spain n.a. 4.6 7.2 6.2 7.2 4.1 4.8 4.5 6.3 5.6 3.2 -1.4 -0.5France n.a. 3.2 8.8 6.7 6.1 5.8 3.7 2.9 9.5 6 6.7 -0.9 -0.3Italy 3) n.a. 4.6 5.8 6.2 4.5 4.2 5.2 1.6 3.9 4.8 4.6 -0.6 -0.2Luxembourg n.a. 0.6 3.9 2.9 5.5 2.4 6.2 2.4 0.0 2.8 6.2 -0.2 -0.1Netherlands n.a. 1.5 3.8 2.1 2.9 2.4 1.6 2.1 2.6 2.5 2.4 -1.1 -0.4Austria n.a. 2.0 4.3 3.2 2.6 2.1 1.1 1.5 2.8 2.5 3.3 -2.3 -1.2Portugal n.a. 4.8 8.3 5.3 5.8 5.9 1.7 1.8 10.9 5.4 9.1 4.7 2.3Slovenia n.a. 1.4 7.1 3.9 1.8 1.9 2.4 2.5 4.3 3 5.7 -2.0 -0.7Finland n.a. 1.5 6.6 4.0 7.0 2.6 3.7 1.8 2.3 3.3 5.5 0.6 0.2Euro area 6) 2.2 3.5 6.5 5.4 5.6 4.6 3.9 3.1 5.4 4.9 3.4 -0.9 -0.3Denmark n.a. 2.1 6.8 3.7 4.5 2.6 1.9 2.4 3.1 3.2 4.9 -8.0 -2.7 Sweden n.a. 2.6 7.9 5.3 6.6 4.9 3.7 2.2 2.8 4.2 5.7 -2.3 -1.1 United
Kingdom n.a. 1.4 3.0 2.3 2.9 2.0 1.2 1.3 4.6 2.2 3.4 0.4 0.2
Sources: EU-LFS (spring data) and ECB calculations. Notes: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. 1) 25 to 64 years old. 2) Differences in the total unemployment rate compared to table 4.1.a are due to missing data. 3) Starting 2004 for AT, BE, IE, PT, SE, UK. 4) Data for Ireland is 2005. 5) Average annual changes are presented in yellow. 6) EA is calculated without IE. The category "general" is excluded from this analysis due to questionable data quality.
104ECB
Occasional Paper No 87
June 2008
Table 35 Unemployment rates and mismatch by type of education (15-29 year-olds only) in the euro area and the euro area countries 1)
Unemployment rate in 2006 (%) Mismatch (range)
Country General
Teacher training
and educa-
tion
Humani-ties
langu-ages and
art
Social science
business, law
Science math
compu-ting
Engine-ering
manufac-turing
Agri-culture
Health and
welfareSer -
vices
Weighted average of
sub- groups 2) 2006
Change (p.p.) 5)
2004-2006 3)
Belgium 15.2 7.6 18.6 13.2 11.7 8.8 6.9 9.6 14.8 14.6 11.8 -5.6 -2.8Germany 5.4 5.0 10.7 10.4 5.7 11.8 13.2 6.7 11.9 11.8 8.2 -3.2 -1.1Ireland 4) 5.2 2.4 5.6 2.3 5.3 3.8 1.3 3.4 3.0 5.9 4.3 n.a. n.a.Greece 17.0 25.5 24.7 20.3 22.4 14.3 27.7 27.4 18.5 17.8 13.4 1.7 0.6Spain 16.2 11.5 12.4 10.4 14.4 7.8 14.2 11.0 13.3 14.1 8.4 -3.6 -1.2France 19.0 0.0 17.4 14.5 8.8 10.1 10.1 9.0 16.3 16.2 19.0 4.7 1.6Italy 24.4 17.0 18.3 15.8 19.8 11.0 12.4 13.5 16.5 15.8 13.3 -15.3 -5.1Luxembourg 15.4 6.1 11.0 7.5 17.4 5.8 0.0 2.9 6.0 10.4 17.4 6.9 2.3Netherlands 7.0 3.3 7.1 3.7 5.2 3.5 5.4 3.8 3.3 6.4 3.9 -2.4 -0.8Austria 6.8 4.2 6.9 6.1 2.9 4.6 5.2 2.3 8.3 7.5 6.0 0.9 0.5Portugal 12.3 19.0 12.9 12.7 15.6 9.7 7.0 13.1 13.3 13.3 12.1 2.7 1.3Slovenia 11.1 5.1 19.8 13.2 6.6 9.2 17.0 10.0 11.6 11.9 14.6 1.9 0.6Finland 18.0 0.0 18.2 7.1 15.8 8.9 12.4 4.8 9.5 18.4 18.2 -0.2 -0.1Euro area 6) 11.8 9.8 15.4 12.1 12.3 10.0 11.1 8.5 12.8 13.6 6.9 -1.9 -0.6Denmark 7.7 3.3 6.3 6.1 6.3 3.8 4.2 5.0 5.7 6.7 4.4 -8.1 -2.7Sweden 12.8 7.8 13.7 10.4 7.1 9.9 12.3 10.5 9.1 15.4 6.6 -3.5 -1.7United
Kingdom 9.7 3.6 5.5 5.9 8.5 5.1 5.6 5.7 8.2 10.4 6.1 1.7 0.9
Sources: EU-LFS (spring data) and ECB calculations. Notes: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. The detailed education breakdown is only available for 2006. 1) 15 to 29 years old. 2) Differences in the total unemployment rate to table 4.1.a are due to missing data. 3) Starting 2004 for AT, BE, IE, PT, SE, UK. 4) Data for Ireland is 2005. 5) Average annual changes are presented in yellow. 6) EA calculated without IE.
105ECB
Occasional Paper No 87
June 2008
ANNEX 3
Table 36 Unemployment rates and mismatch by level of educational attainment in the euro area and the euro area countries (15-29 year olds only) 1)
Country
Unemployment rate in 2007 (%) Educational mismatch
Low Medium High
Weighted average of subgroups
Range 2007
Change in range between highest and lowest rate (p.p)
1993-1995 1996-2001 2002-2007Belgium 23.9 14.2 7.5 13.9 16.4 8.4 2.8 0.2 0.0 -1.3 -0.2Germany 18.2 8.2 4.7 10.9 13.6 3.1 1.0 0.8 0.1 6.6 1.1Ireland 15.0 6.6 3.3 6.7 11.7 1.5 0.5 -12.2 7) -2.0 7) 3.47) 0.6 7)
Greece 14.3 17.3 19.2 17.0 4.9 -1.2 -0.4 -4.6 -0.8 1.4 0.2Spain 15.8 12.4 8.0 12.6 7.8 -4.2 -1.4 0.0 0.0 5.1 0.8France 4) 29.7 13.4 8.1 15.0 21.5 n.a. n.a 1.5 0.2 3.1 0.5Italy 15.5 12.1 14.4 13.4 3.4 3.0 1.0 -6.4 -1.1 0.9 0.1Luxembourg 15.2 7) n.a. n.a. 7.5 n.a. n.a. n.a n.a. n.a. n.a. n.a.Netherlands 5) 8.4 2.6 1.5 7) 4.5 6.9 7) n.a. n.a -5.3 -0.9 3.5 7) 0.6 7)
Austria 3) 11.5 5.4 n.a. 6.8 n.a. n.a. n.a 0.1 0.0 n.a. n.a.Portugal 13.1 12.2 14.0 13.0 1.8 -0.2 7) -0.1 7) -2.0 7) -0.3 7) 0.1 7) 0.0 7)
Slovenia 5) 13.1 7.3 5.8 7.6 7.2 n.a. n.a 5.1 7) 0.9 7) -8.4 7) -1.4 7)
Finland 3) 29.8 11.3 4.1 7) 15.7 25.7 7) n.a. n.a -4.8 -0.8 -4.4 7) -0.7 7)
Euro area 17.6 10.3 7.9 12.0 9.7 2.4 0.8 -1.6 -0.3 2.7 0.5Denmark 7.9 4.1 4.5 7) 5.8 3.8 -5.3 7) -1.7 7) 0.9 7) 0.2 7) 0.0 7) 0.0 7)
Sweden 3) 33.9 11.4 7.7 16.6 26.2 n.a. n.a -1.7 -0.3 12.3 2.1United
kingdom 21.6 9.2 3.6 10.4 18.0 -0.1 0.0 2.0 0.3 5.1 0.9
Sources: EU-LFS (spring data) and ECB calculations. Note: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. 1) The education levels refer to low: lower secondary education and less, medium: upper secondary education, high: tertiary education. In bold are the three best performers in terms of a low level and low range of unemployment rates.2) Education data start in 1992. 3) Data start in 1995. 4) Data start in 1993. 5) Data start in 1996. 6) Average annual changes are presented in the smaller font in yellow. 7) Based on fi gures smaller than the Eurostat reliability limit.
106ECB
Occasional Paper No 87
June 2008
Table 37 Unemployment rates and mismatch by level of educational attainment in the euro area and the euro area countries (non-nationals, 25 to 64 year olds only) 1)
Unemployment rate in 2007 (%) Educational mismatch
Country Low Medium High
Weighted average of subgroups
Range2007
Change in range between highest andlowest rate (p.p) 7)
1996-2001 2002-2007
Belgium 25.2 14.6 7.9 16.0 17.3 -5.7 -0.9 6.0 1.0 Germany 21.1 13.4 11.3 16.2 9.7 1.3 0.2 4.8 0.8 Ireland 2) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.Greece 6.8 9.7 7.7 8.0 3.0 n.a. n.a. 0.7 0.1 Spain 13.9 10.4 10.7 12.0 3.6 -0.3 8) 0.0 8) -3.6 -0.6 France 20.6 15.2 n.a. 17.2 n.a. 0.6 0.1 n.a. n.a. Italy 3) 8.8 6.3 7.2 7.6 2.5 n.a. n.a. n.a. n.a. Luxembourg 4.6 n.a. 4.0 4.0 n.a. -1.3 -0.2 n.a. n.a. Netherlands 4) 11.3 5.4 6.1 6.8 5.9 n.a. n.a. n.a. n.a. Austria 16.4 8.7 n.a. 10.8 n.a. 1.0 0.2 n.a. n.a. Portugal 4) 18.3 n.a. n.a. 14.0 n.a. n.a. n.a. n.a. n.a. Slovenia 5) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Finland 29.2 14.3 10.4 17.9 18.8 n.a. n.a. n.a. n.a. Euro area 16.5 11.1 10.0 13.2 6.5 2.0 0.3 0.3 0.0 Denmark 14.5 n.a. n.a. 9.9 n.a. n.a. n.a. n.a. n.a. Sweden 6) 22.5 10.7 10.5 13.4 12.0 n.a. n.a. 5.5 0.9 United Kingdom 12.4 8.0 5.3 7.9 7.1 2.5 0.4 -2.1 -0.4
Sources: EU-LFS (spring data) and ECB calculations. Note: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. 1) Dataset starts 1995; the education levels refer to low: lower secondary education and less, medium: upper secondary education, high: tertiary education. 2) Only data for 1998-2004. 3) Only data for 2005-07. 4) Data start 1999. 5) Data start 2002. 6) Data start 1997. 7) Average annual changes are presented in yellow.8) Based on fi gures smaller than the Eurostat reliability limit.
107ECB
Occasional Paper No 87
June 2008
ANNEX 3
Table 38 Unemployment rates and mismatch by level of educational attainment in the euro area and the euro area countries (non-nationals aged 15-29 years old) 1)
Country
Unemployment rate in 2007 (%) Educational mismatch
Lower secondary education
and less
Upper secondary education
Tertiaryeducation
Weightedaverage ofsubgroups
Range2007
Change in range between highest and lowest rate (p.p) 7)
1996-2001 2002-2007
Belgium 36.2 23.7 8.8 24.5 27.4 2.1 0.3 12.4 2.1 Germany 24.9 12.8 9.9 17.8 15.0 0.6 0.1 7.7 1.3 Ireland 2) n.a. n.a. n.a. n.a. n.a. n.a. n.a n.a. n.a Greece 6.4 17.8 12.3 10.4 11.3 -3.1 -0.5 6.5 1.1 Spain 17.0 12.5 8.6 14.1 8.4 -12.6 -2.1 -0.1 0.0 France 32.7 19.2 15.9 24.4 16.7 3.1 0.5 -0.7 -0.1Italy 3) 10.3 9.3 12.2 10.1 2.8 n.a. n.a n.a. n.a Luxembourg 12.8 2.1 3.1 6.3 10.8 1.1 0.2 8.2 1.4 Netherlands 10.8 8.3 5.5 8.6 5.4 n.a. n.a -0.8 -0.1Austria 20.7 10.8 11.4 14.8 9.9 -7.2 -1.2 5.8 1.0 Portugal 4) 25.9 10.5 12.2 18.6 15.4 n.a. n.a -4.8 -0.8Slovenia 5) 0.0 0.0 n.a. 0.0 0.0 n.a. n.a n.a. n.a Finland 42.0 25.2 0.0 31.3 42.0 1.2 0.2 -2.4 -0.4Euro area 19.7 12.9 10.2 15.7 9.5 1.4 0.2 1.6 0.3 Denmark 19.0 10.5 11.9 13.2 8.5 47.9 8.0 -41.9 -7.0Sweden 6) 42.9 13.8 12.6 24.3 30.3 n.a. n.a 10.6 1.8 United Kingdom 15.4 10.5 7.2 10.4 8.3 1.9 0.3 -10.0 -1.7
Sources: EU-LFS (spring data) and ECB calculations. Note: For all tables on mismatch, the fi gures presented are those fulfi lling Eurostat's publications and reliability limits for the LFS data. 1) 15 to 29 years old; dataset starts 1995; the education levels refer to low: lower secondary education and less, medium: upper secondary education, high: tertiary education. 2) Only data for 1998-2004. 3) Only data for 2005-07. 4) Data start in 1999.5) Data start in 2002.6) Data start in 1997. 7) In yellow are average annual changes, a negative sign indicates that mismatch has decreased.
108ECB
Occasional Paper No 87
June 2008
ANNEX 4
Box 12
CROSS-BORDER COMMUTING IN THE EURO AREA. CASE STUDY: LUXEMBOURG1
The number of outward-commuters from the euro area countries reached more than 2 million
persons in 2006 and has more than tripled over the past ten years.2 Commuters currently
represent 1.6% of euro area employment, with 87% of all commuters coming from only three
countries, namely Italy, Germany and France3. Outward-commuters are essentially men (83%)
aged between 25 and 54 (81%) with generally a low or medium level of education (respectively
43% and 41%). Over the last ten years, the share of cross-border commuters from euro area
countries with a “high” level of education (relative to those residing and working in the same
country) has remained relatively low.
Recent work analysing the pull and push factors on regional commuting fl ows in the European
Union 4 shows that commuting is well explained by the standard explanatory factors such as
the size of origin and destination regions and wage differentials. More specifi cally, high
unemployment and low wages in the home regions push workers towards commuting. For the
host country, cross-border labour supply results in a more effi cient allocation of labour across the
internal market. For the country of origin, commuters act as a buffer because outward commuting
reduces domestic unemployment in the case of asymmetric shocks. Moreover, commuting, as an
alternative to unemployment, allows workers to keep or even improve their skills, while reducing
unemployment benefi t-related expenditures. The home country also benefi ts from commuters’
incomes, which stimulate consumption.
Cross-border commuting in Luxembourg
Luxembourg does not have a continuous history of immigration, but has rather experienced three
distinct immigration waves: (i) Italians from the late nineteenth century to the 1950s, (ii) Portuguese
in the 1960s and 1970s - both characterised by a tendency towards permanent migration – and
(iii) since the 1980s, immigrants from a larger number of countries. The most recent estimates of
migration fl ows show Luxembourg as having the highest proportion of foreigner residents in its
population. In 2006, approximately 40% of the resident population were non-nationals. The more
recent phenomenon of a very rapidly growing number of commuters (the “frontaliers”) is closely
linked to opportunities in the labour market. In April 2007 about 39% of total employment and
68% of new jobs were occupied by cross-border workers.5 The expansion of the service industries,
notably fi nancial services and media-related companies during the last 20 years, has continued the
tendency for cross-border workers to be disproportionately over-represented in both low-skilled,
manual manufacturing jobs (55.5% of total employees in that sector) and construction (47.8%), in
high-skilled jobs in the fi nancial sector (48.4%) and in real estate, renting and business activities
1 Prepared by C. Olsommer.
2 Figures on outward commuting fl ows in the euro area are taken from Eurostat.
3 With respectively 1,373,284; 289,647 and 229,392 outward commuters in 2006.
4 J. Marvakov and T. Mathä (2006).
5 These high numbers are partly a refl ection of Luxembourg’s small size and geographical location in the middle of a large economic
area. The migration over such small distances would constitute only internal migration in Luxembourg’s larger neighbours. 51% of
cross-border immigrants are from France, 26% from Germany and 23% from Belgium.
109ECB
Occasional Paper No 87
June 2008
ANNEX 4
(57.2%). Average educational attainment differs considerably among the various nationalities of
foreign residents and “frontaliers” working in Luxembourg.
The most recent studies about the determinants of cross-border migration in Luxembourg
emphasise the neo-classical theory that individuals’ migration decisions are determined partly
by the expected income from work. Average monthly wages and the minimum wage levels in
Luxembourg are higher than in the neighbouring countries.6 This wage differential between the
expected wage in Luxembourg and in the residence country might be especially important for
commuters, since they are likely to spend most of their wage in their country of residence.7 Other
factors infl uencing the cross-border commuting decision include the fi nancial and non-fi nancial
costs of such temporary migration. Lastly, a high probability of fi nding a job,8 combined with
access to social security benefi ts (in particular child benefi ts and relatively generous parental
leave) increase the advantages relative to costs of temporary migration to Luxembourg.
These migrant fl ows have an impact on labour supply in both Luxembourg and its neighbouring
countries. Luxembourg experiences a shortage of skilled labour, with more than a half of its
resident unemployed being low skilled.9 Temporary immigration helps to reduce labour shortages,
particularly in the fi nancial sector by supplying specialised and high-skilled labour. Commuters
therefore complement labour supply from the resident labour force and facilitate domestic factor
utilisation. Neighbouring countries also benefi t from this temporary migration, since jobs seekers,
by fi nding a job in Luxembourg, reduce the unemployment rate in their country of residence.10
Luxembourg’s experience, despite some traffi c problems, provides a successful example of
cross-border labour mobility helping to complement the domestic labour force.
6 OECD annual data show the monthly minimum wage (in PPS) in 2005 to be: €1,417 for unskilled workers and €1,700 for skilled
workers in Luxembourg, €1,218 in France and €1,234 in Belgium.
7 Although, of course, this wage differential may be more prominent in certain sectors and thus more relevant for particular groups of
potential commuters.
8 Job creation was vigorous in Luxembourg over the past twenty years, reaching on average 3.6% per year. The cross-border workers
largely benefi ted from this, as they occupied 7/10ths of the new jobs on average since 1986.
9 Luxembourg’s Beveridge curve shifted out during the 1990s and has done so again since 2005, indicating an increasing degree of
mismatch between supply and demand. This conclusion is also supported by alternative measures of structural change, such as the
Lilien indicator of inter-sectoral structural change and the rate of unsatisfi ed sectoral labour demand over unsatisfi ed sectoral labour
supply. See also Banque centrale du Luxembourg (2004).
10 All the more since the unemployment rates in neighbouring states are higher than in Luxembourg (4.5% in 2006). Lorraine (France):
11.0%, Saarland (Germany): 9.7%, Rhineland-Palatinate (Germany): 8.2% and Wallonia (Belgium): 18.8%.
110ECB
Occasional Paper No 87
June 2008
REFERENCES
Aaronson, S., B. Fallick, A. Figura, J. Pingle and W. Wascher (2006), “The Recent Decline in the
Labour Force Participation Rate and Its Implications for Potential Labour Supply”, Brookings
Papers on Economic Activity, March.
Acemoglu, D. and J.S. Pischke (1998), “Why do Firms Train? Theory and Evidence”, Quarterly
Journal of Economics, Vol. 113:1, pp. 79-119.
Acemoglu, D. and J.S. Pischke (1999), “Beyond Becker: Training in Imperfect Labour Markets”,
Economic Journal, Vol. 109:453, pp. F112-F142.
Acemoglu, D., D.H. Autor and D. Lyle (2002), “Women, War and Wages: The Effect of Female
Labor Supply on the Wage Structure at Mid-century”, NBER Working Papers, No 9013.
Acemoglu, D. (2003), “Cross-Country Inequality Trends”, Economic Journal, February 2003, Vol 113,
pp. F121-149.
Aghion, P, M. Dewatripont, C. Hoxby, A. Mas-Colell and A. Sapir (2007), “Why reform Europe’s
universities?”, Bruegel policy brief, September 2007/04.
Amable, B., L. Demmou and D. Gatti (2007), “Employment Performance and Institutions:
New Answers to an Old Question”, IZA Discussion Paper, No 2731.
Ammermüller, A. and A.M. Weber (2005), “Educational Attainment and Returns to Education
in Germany”, ZEW-Discussion Paper, No 05-17.
Amuedo-Dorantes, C. and S. de la Rica (2007), “Labour Market Assimilation of Recent Immigrants
in Spain”, forthcoming in the British Journal of Industrial Relations.
Antunes, A. and T. Cavalcanti (2007), “Start Up Costs, Limited Enforcement, and the Hidden
Economy”, European Economic Review 51(1), pp. 203-24.
Ashenfelter, O., C. Harmon and H. Oosterbeek (1999), “A Review of Estimates of the
Schooling/Earnings Relationship, with Tests for Publication Bias”, Labour Economics 6(1),
pp. 453-470.
Baker, M. and D. Benjamin (1994), “The Performance of Immigrants in the Canadian Labour
Market”, Journal of Labour Economics, Vol. 12, pp.369-405.
Bank of Italy (2006), Annual Report.
Banque centrale du Luxembourg (2004) “(Le) Chômage structurel (encadré)”, in Bulletin de
la Banque centrale du Luxembourg, n° 4/2004, pp.36-40.
Barrett, A., A. Bergin and D. Duffy (2006), “The Labour Market Characteristics and Labour Market
Impacts of Immigrants in Ireland”, IZA Discussion Paper, No 1553 and Economic and Social
Review Vol.37, No 1.
111ECB
Occasional Paper No 87
June 2008
REFERENCES
Barrett, A. and Y. McCarthy (2007), “Immigrants in a Booming Economy: Analysing their Earnings
and Welfare Dependence”, Labour: Review of Labour Economics and Industrial Relations,
Vol. 21 (4-5), pp. 789-808.
Barro R. (1991), “Economic growth in a cross section of countries”, Quarterly Journal of Economics,
May, 106(2), pp. 407-43.
Barro R. and X. Sala-I-Martin (1995), “Capital mobility in neoclassical models of growth”,
American Economic Review, 85, pp. 103-115.
Barro R. (2001), “Human capital and growth”, American Economic Review, May, 91(2), pp.12-17.
Barro R. and X. Sala-I-Martin (2004), “Economic Growth”, MIT Press, Cambridge.
Bassanini, A. and R. Duval (2006), “Employment Patterns in OECD Countries: Reassessing the
Role of Policies and Institutions”, OECD Social, Employment and Migration Working Papers
No 35, Paris.
Becker, G. (1962), “Investment in human capital: a theoretical analysis”, Journal of Political
Economy, Vol. 70:5, pp. 9-49.
Benhabib J. and M.M. Spiegel (1994), “The role of human capital in economic development:
Evidence from aggregate cross-country data”, Journal of Monetary Economics, October, 34(2),
pp. 143-74.
Berger, M.C. (1985), “The Effect of Cohort Size on Earnings Growth: A Reexamination of the
Evidence”, The Journal of Political Economy, Vol. 93, No 3, pp. 561-573.
Bertola, G., F. Blau and L. Kahn (2002), “Labor Market Institutions and Demographic Employment
Patterns”, NBER Working Paper, No 9043.
Blanchard, O. and J. Wolfers (2000), “The Role of Shocks and Institutions in the Rise of European
Unemployment: The Aggregate Evidence”, The Economic Journal, Vol. 110, March.
Blanchard and Gali (2007), “Real wage rigidities and the New Keynesian Model”, Journal
of Money, Credit and Banking, Vol. 39:2, pp.35-66
Blöndal, S. and S. Scarpetta (1999), “The Retirement Decision in OECD Countries”, OECD
Working Paper, No 202.
Blundell, R. and T. Macurdy (1999), “Labor Supply: A Review of Alternative Approaches”,
in Ashenfelter, O. and Card. D. (eds), Handbook of Labor Economics 3, pp. 1559-1695.
Borjas, G. (2000), “Foreign-Born Teaching Assistants and the Academic Performance
of Undergraduates,” American Economic Review, Vol. 90(2), pp. 355-359.
Borjas, G. (2001), “Heaven’s Door”, Princeton University Press.
112ECB
Occasional Paper No 87
June 2008
Bosworth, B.P. and S.M. Collins (2003), “The empirics of growth: An update”, Brookings Papers
on Economic Activity 2003, No 2, pp. 113-206.
Bradbury K. (2005), “Additional slack in the economy: the poor recovery in labor force participation
during this business cycle”, Federal Reserve Bank of Boston, No 05-2.
Brandolini, A. and P. Cipolline (2001), “Multifactor Productivity and Labour Quality in Italy,
1981-2000”, Temi di discussione, Banca d’Italia.
Brandolini, A, P. Cipollone and E. Viviano (2006), “Does ILO Defi nition Capture all
Unemployment?”, Journal of the European Economic Association, 4(1), pp. 153-179.
Budría, S and P.T. Pereira (2005), “Educational Qualifi cations and Wage Inequality: Evidence for
Europe,” IZA Discussion Paper, No 1763.
Burda, M. D. Hamermesh and P. Weil (2006), “The Distribution of Total Work in the EU and US”,
IZA Discussion Paper, No 2270.
Caliendo, M., A. Kritikos, V. Steiner and F. Wießner (2007), “Existenzgründungen: Unterm Strich
ein Erfolg”, IAB Kurzbericht 10, IAB.
Callan, T., A. van Soest and J. Walsh (2003), “Tax cuts, tax reform and labour supply responses”,
Study for the Social Economic Research Institute, Dublin.
Callan, T., A. van Soest and J.R. Walsh (2007), “Tax Structure and Female Labour Market
Participation: Evidence from Ireland,” IZA Discussion Paper, No 3090.
Card, D. and R. Freeman (2004), “What Have Two Decades of British Economic Reform
Delivered?”, in Card, D., R. Blundell and R. Freeman (editors): Seeking a Premier Economy:
The Economic Effects of British Economic Reforms, 1980-2000, NBER.
Carnoy, M., M. Castells and C. Brenner (1997), “Labour Markets and Employment Practices in the
Age of Flexibility”, International Labour Review, Vol.136, Spring.
Carone, G., and A. Salomäki (2001), “Reforms in Tax-Benefi t Systems in order to Increase
Employment Incentives in the EU”, DG-ECFIN Economic Paper 160.
Carrasco, R., J.F Jimeno and A.C. Ortega (2008), “The effects of immigration on the labour market
performance of native-born workers: some evidence for Spain”, forthcoming in Journal of
Population Economics, Vol.3/2008.
Centeno, M. and P. Fernandes (2004), “Labour Market Heterogeneity: Distinguishing between
Unemployment and Inactivity”, Banco de Portugal, Economic Bulletin.
Centeno, M. and A. Novo (2007), “Identifying Unemployment Insurance Income Effects with a
Quasi-Natural Experiment”, Banco de Portugal Working Paper, No 10.
113ECB
Occasional Paper No 87
June 2008
REFERENCES
Chiswick, B. R. and P.W. Miller (1995), “The Endogeneity between Language and Earnings:
International Analyses,” Journal of Labor Economics, Vol. 13(2), pp. 246-288.
Chiswick, B. R. and P.W Miller (2005), “Why is the Payoff to schooling smaller for immigrants?”,
IZA Discussion Paper, No 1731.
Christoffel, K.P., K. Kuester and T. Linzert (2006), “Identifying the role of labor markets for monetary
policy in an estimated DSGE model”, Working Paper Series, No 635, European Central Bank.
Ciccone, A., and E. Papaioannou (2005), “Human Capital, the Structure of Production, and
Growth”, Barcelona, Universitat Pompeu Fabra.
Clark K.B. et al. (1979), “Labor Market Dynamics and Unemployment: A Reconsideration”,
Brookings Papers on Economic Activity, No 1, pp.13-72.
CNAF (2005), “La PAJE après 18 mois de montée en charge”, L’e-ssentiel, n°42.
Copenhagen Economics (2007) “Study on reduced VAT applied to goods and services in the
Member States of the European Union”, Kopenhagen. www.ec.europa.eu/taxation_customs/
resources/documents/taxation/vat/how_vat_works/rates/study_reduced_VAT.pdf
Coulombe, S., J.F. Tremblay and S. Marchand (2004), “Literacy scores, human capital and growth
across fourteen OECD countries”, Statistics Canada, Ottawa.
Coulombe, S. and J.F. Tremblay (2006), “Literacy and Growth”, Topics in Macroeconomics 6, No 2.
Daly, Kevin (2007), “Gender inequality, growth and global ageing”, Goldman Sachs Global
economics paper, No. 154.
Deutsche Bundesbank (2006), “The evolution of labour market related-expenditure”, Monthly
Bulletin, September.
Duval, R. (2003), “The retirement effects of old-age pension and early retirement schemes in OECD
countries”, OECD Working Paper, No 304.
Duval, R. (2006), “Pension Systems, Social Transfer Programmes and the Retirement Decision
in OECD Countries”, in Messina, J., C. Michelacci, J. Turunen G. and Zoega (eds.)
Labour Market Adjustments in Europe, Edward Elgar, Cheltenham.
Duval R., J. Elmeskov and L. Vogel (2007), “Structural Policies and Economic Resilience to
Shocks”, OECD Economics Department Working Papers, No 567, OECD.
Easterlin, R. (1980), “Birth and Fortune: The Effects of Generation Size on Personal Welfare”,
New York: Basic Books.
Eichhorst, W. and K.F. Zimmermann (2007), “How many (and Which) Measures of Active Labor
Market Policy Do We Still Need?”, IZA Discussion Paper, No 2605.
114ECB
Occasional Paper No 87
June 2008
Elmeskov J. and K. Pichelmann (1993), “Interpreting unemployment: the role of labour-force
participation”, OECD Economic Studies, No 21.
European Central Bank (2002), “Labour Market Mismatches in Euro Area Countries”, March,
Frankfurt.
European Central Bank (2006), “Developments in the structural features of euro area labour markets
over the last decade”, Monthly Bulletin, September.
European Commission (2002a), “Social Situation Report 2002”, Brussels.
European Commission (2002b), “The Gender Pay Gap – Origins and policy responses:
A comparative review of 30 European countries”, Paper by a group of experts on Gender, Social
Inclusion and Employment, July 2006.
European Commission (2006), “The impact of ageing on public expenditure: Projections for the
EU25 Member States on pensions, health-care, long-term care, education and unemployment
transfers (2004-2050)”, European Economy, Special Report No 1.
Employment Committee (2007), “Employment Committee Working Group Report on Active
Ageing”, http://ec.europa.eu/employment_social/employment_strategy/emco/activeageing_
workinggroup07_en.pdf.
Eurostat (1996), “The European Union Labour Force Survey: Methods and Defi nitions”, Offi ce for
Offi cial Publications of the European Communities, Luxembourg.
Eurostat (2007), “Reconciliation between work and family life”.
Fallick, B. and J. Pingle (2007), “A Cohort Based Model for Labour Force Participation”, Finance
and Economics Discussion Series, Board of Governors of the Federal Reserve System.
Fernández, C. and A.C. Ortega (2006), “Labour Market Assimilation of Immigrants in Spain:
Employment at the Expense of Bad Job-Matches?”, IESE Business School working paper, No 644.
Finegold, D and D. Soskice (1988), “The failure of training in Britain: analysis and prescription”,
Oxford Review of Economic Policy, Vol.4:3, pp.21-53.
Fiori, G., G. Nicoletti, S. Scarpetta, and F. Schiantarella (2007), “Employment Outcomes and
the Interaction Between Product and Labor Market Deregulation: Are They Substitutes or
Complements?”, IZA Discussion Paper, No 2770.
Freeman, R.B. and R. Schettkat (2005), “Marketization of Household Production and the
EU-US Gap in Work”, Economic Policy, pages 5-50.Genre, V., R. Gomez-Salvador and A. Lamo
(2005b), “The Determinants of Labour Force Participation in the EU” in R. Gómez-Salvador,
A. Lamo, B. Petrongolo, M. Ward, and E. Wasmer (eds.) Labour Supply and Incentives to Work
in Europe, Edward Elgar.
115ECB
Occasional Paper No 87
June 2008
REFERENCES
Genre, V. and R. Gomez-Salvador (2002), “Labour force developments in the euro area since the
1980s” ECB Occasional Paper, No 4.
Genre, V., R. Gomez-Salvador and A. Lamo (2007), “European Women. Why do(n’t) they work?”,
Applied Economics, forthcoming.
Gomez-Salvador, R., A. Musso, M. Stocker and J. Turunen (2006), “Labour productivity
developments in the euro area”, ECB Occasional Paper, No 53.
González L. (2005), “Nonparametric Bounds on the Returns to Language Skills”, Journal of Applied
Econometrics, Vol 20(6), pp. 771-795.
Gronau, R. and Y. Weiss (1981), “Expected Interruptions in Labour Force Participation and
Sex-Related Differences in Earnings Growth”, Review of Economic Studies, Vol. 48(4),
pp. 607-19.
Gustafsson, S. and F.P. Stafford (1992), “Child care subsidies and Labour supply in Sweden”,
Journal of Human Resources, Vol. 27, no.1, pp. 204-230.
Hanushek E. and D. Kimko (2000), “Schooling, labour-force quality, and the growth of nations”,
American Economic Review, December, 90(5), pp. 1184-1208.
Hanushek, E and L. Wößmann (2007), “The role of education quality for economic growth”, World
Bank Policy Research, Working Paper, No 4122, Washington, World Bank.
Inklaar, R., M. O’Mahony and M. Timmer (2005), “ICT and Europe’s Productivity Performance:
Industry Level Growth Account Comparisons with the United States”, Review of Income and
Wealth 51(4), pp. 505-536.
International labour review (1997), International Labour Organization.
Jacobi, L. and J. Kluve (2007), “Before and after the Hartz reforms: The performance of active
labour market policy in Germany”, Zeitschrift für Arbeitsmarktforschung, 40, 1, pp. 45-64.
Jaumotte, F. (2003), “Female Labour Force Participation: Past Trends and Main Determinants in
OECD Countries”, OECD Economics Department Working Papers No 376.
Jimeno, J.F. and D. Rodriguez-Palenzuela (2002), “Youth unemployment in the OECD:
demographic shifts; labour market institutions, and macroeconomic shock”, Working Paper,
No 155, European Central Bank.
Jorgenson, D. (2005), “Accounting for Growth in the Information Age”, in P. Aghion and
S. Durlauf, eds., Handbook of Growth Economics, Vol 1A, pp.743-815.
Jones, S.R.G. and W.C. Riddell (1999), “The Measurement of Unemployment: An Empirical
Approach”, Econometrica, 67, pp. 147-162.
Kailis, E. and S. Pilos (2005), “Lifelong learning in Europe”, Eurostat Statistics in Focus, 8/2005.
116ECB
Occasional Paper No 87
June 2008
Kanellopoulos, K., K. Mavromaras and T. Mitrakos (2004), “Education and the labour market”,
Study No 50, KEPE: Athens (in Greek).
Kate C., H. Bachnan and F. J. Morrison (2001), “Schooling and Cognitive Development,” in
Sternberg, R.J. and E.L. Grigorenko (eds.) Environmental Effects on Cognitive Abilities,
Mahwah, New Jersey: Lawrence Erlbaum Associates, Publishers, Chapter 10, pp. 287 - 335.
Kettner, A. and M. Rebien (2007), “Hartz-IV-Reform: Impulse für den Arbeitsmarkt“,
IAB Kurzbericht 19, IAB.
Killingsworth, M. and J. Heckman (1986), “Female Labor Supply: A Survey”, in: Ashenfelter,
O. and Layard, R. (eds), Handbook of Labor Economics (1), North-Holland, Amsterdam,
pp. 103-204.
Korenman, S. and D. Neumark (1997), “Cohort Crowding and Youth Labor Markets: A
Cross-National Analysis”, NBER Working Paper, No 6031.
Kossoudji, S.A. and D. Cobb Clark (2002), “Coming out of the Shadows: Learning about Legal
Status and Wages from the Legalized Population”, Journal of Labor Economics, Vol. 20(3),
pp. 598-628.
Lauer, C and V. Steiner (2000), “Private Erträge von Bildungsinvestitionen in Deutschland”,
Beihefte der Konjunkturpolitik, Applied Economics Quarterly, 51, pp. 71-101.
Layard R., S. Nickell and R. Jackman (2005), ”Unemployment”, Chapter 8, Oxford University
Press.
Leiner-Killinger N., C. Madaschi and M. Ward-Warmedinger (2005), “Trends and patterns in
working time across euro area countries 1970-2004: causes and consequences”, ECB Occasional
Paper No 41, December 2005.
Levenson, A. and C. Zoghi (2007), “The Strength of Occupation Indicators as a Proxy for Skill”
Working Papers 404, U.S. Bureau of Labor Statistics.
Lilien, D. and R.E. Hall (1986), “Cyclical Fluctuations in the Labor Market”, in 0. C.Ashenfelter
and R. Layard, eds., Handbook of Labor Economics, Vol. 2, Amsterdam: North-Holland.
Lipsey, R.G. (1960), “The Relation between Unemployment and the Rate of Change of Money
Wage Rates in the United Kingdom, 1862-1957: A Further Analysis”, Economica, Vol. 27,
No 105. (Feb., 1960), pp. 1-31.
Lopez-Garcia, P. (2003) “Labour Market Performance and Start-up Costs: OECD evidence”,
mineo.
Maddaloni, A., A. Musso, P. Rother, M. Ward-Warmedinger and T. Westermann (2006),
“Macroeconomic Implications of Demographic Developments in the Euro Area”, European
Central Bank Occasional Paper, No 51.
117ECB
Occasional Paper No 87
June 2008
REFERENCES
Magoula, T. and G. Psacharopoulos (1999), “Schooling and monetary rewards in Greece: an
over-education false alarm?”, Applied Economics 31: 12, pp.1589-97.
Martins, P., A. Novo and P. Portugal (2007), “Increasing the Legal Retirement Age: The Impact on
Wages, Hours, Worker Flows and Firm Performance”, Banco de Portugal, mimeo.
Marvakov, J. and T.Y. Mathä (2007), “An analysis of regional commuting fl ows in the European
Union”, Cahier d’Etudes n°28, Banque Centrale du Luxembourg.
Melka, J. and L. Nayman (2004), “Labour Quality and Skill Biased Technological Change in
France”, CEPII, unpublished.
Micheaux, S and O. Monso (2007), “Faire garder ses enfants pendant son temps de travail”, Insee
Working Paper, No 1132, April.
Mincer J. and S. Polachek (1974), “Family investments in human capital: earnings of women”,
Journal of Political Economy, Vol. 82 (supplement), pp. S76-S108.
Montanaro P. (2007), “Quality of education and human capital: An overview”, Bank of Italy
discussion paper.
National Research Council, Board on Children and Families, Commission on Behavioral and Social
Sciences and Education (1995), “Immigrant Children and Their Families: Issues for Research
and Policy,” The Future of Children, No 5 (2 Summer/Fall), pp.72 - 89.
Nickell S. and R. Layard (1999), “Labor Market Institutions and Economic Performance”,
in: Ashenfelter, O. and D. Card (eds.), Handbook of Labor Economics (3) North-Holland,
Amsterdam, Chapter 42, pp. 3029-3084.
Nicoletti, G., A. Bassanini, E. Ernst, S. Jean, P. Santiago and P. Swa (2001), “Product and
Labour Markets Interactions in OECD Countries”, OECD Economics Department Working
Paper No 312.
Nicoletti, G. and S. Scarpetta (2005), “Product Market Reforms and Employment in OECD
Countries”, OECD Economics Department Working Paper No 472.
OECD (1998), “Employment Outlook”, Paris.
OECD (2001a), “Employment Outlook”, Paris.
OECD (2001b), OECD Productivity Manual: A Guide to the Measurement of Industry-Level and
Aggregate Productivity Growth, Paris.
OECD (2006a), “Education at a Glance”, Paris.
OECD (2006b), “Employment Outlook”, Paris.
OECD (2006c), “International Migration Outlook”, Sopemi Edition, Paris.
118ECB
Occasional Paper No 87
June 2008
OECD (2006d), “Taxing wages 2005-2006”, Paris.
OECD (2007a), “Babies and Bosses. Reconciling Work and Family Life. A Synthesis of Findings
for OECD Countries”, Paris.
OECD (2007b), “Benefi ts and Wages”, Paris.
OECD (2007c), “Economic Surveys: Austria”, Paris.
OECD (2007d), “Education at a glance”, Paris.
OECD (2007e), “Employment Outlook”, Paris.
OECD (2007f), “International Migration Outlook”, Paris.
OECD (2007g), “Pensions at a glance”, Paris.
OECD (2007h), “Taxing wages 2005-2006”, Paris.
Papapetrou, E. (2007), “Education, labour market and wage differentials in Greece”, Bank of
Greece Economic Bulletin, No 28, pp. 51-73.
Pencavel, J. (1986), “Labor Supply of Men: a Survey”, in: Ashenfelter, O. and R. Layard (eds.),
Handbook of Labor Economics (1), North-Holland, Amsterdam, pp. 3-102.
Petrongolo B. and C. Pissarides (2001), “Looking into the black box: A Survey of the Matching
Function”, 39, pp. 390-431.
Pissarides, C., P. Garibaldi, C. Olivetti, B. Petrolongo and E. Wasmer (2003), “Women in the
Labour Force: How Well is Europe Doing?”, in Boeri, T., C. Pissarides and D. Del Boca (eds.)
Women at Work: an Economic Perspective, Oxford University Press, Oxford.
Rivera-Batiz, F.L. (1999), “Undocumented workers in the labor market: An analysis of the earnings
of legal and illegal Mexican immigrants in the United States”. Journal of Population Economics,
Vol. 12(1), pp. 91-116.
Rogerson, R. (2003), “Structural transformation and the deterioration of European Labour Market
Outcomes”, mimeo, Arizona State University.
Rogerson, R. (2007), “Taxation and market work: is Scandinavia an outlier?” NBER Working
Paper, No 12890.
Romans, F. (2007), “The transition of women and men from work to retirement”, Eurostat Statistics
in Focus 97/2007.
Sala-i-Martin, X. (1997), “I just Ran Four Million Regressions”, Economics Working Papers 201,
Department of Economics and Business, Universitat Pompeu Fabra.
119ECB
Occasional Paper No 87
June 2008
REFERENCES
Schneider. F. (2006), “Shadow economies and corruption all over the world: What do we really
know?”, IZA Discussion Paper, No 2315, Bonn.
Schneider, H., K. Brenke, B. Jesske, L. Kaiser, U. Rinne, M. Schneider, J. Steinwede and
A. Uhlendorff (2007), “Evaluation der Maßnahmen zur Umsetzung der Vorschläge der
Hartz-Kommission: Bericht 2006”, IZA Research Report No 10.
Schwerdt, G. and J. Turunen (2007), “Growth in Euro Area Labour Quality”, Review of Income
and Wealth 53(4), pp. 716-734.
Shimer, R. (1998), “Why is the US unemployment rate so much lower?”, NBER Macroeconomics
Annual, pp. 13-73.
Silva, A. (2005), “Taxes and Labor Supply: Portugal, Europe and the United States”, Banco de
Portugal, mimeo.
Social Security Administration (2006), “Social Security Programs Throughout the World: Europe,
2002”, Offi ce of Research, Evaluation and Statistics, Washington.
Steiner, V. and C. Lauer (2000), “Private Erträge von Bildungsinvestitionen in Deutschland,
Beihefte der Konjunkturpolitik”, Applied Economics Quarterly 51, pp. 71-101.
Steiner V. and K. Wrohlich (2006), “Introducing Family Tax-Splitting in Germany: How Would it
Affect the Income Distribution and Work Incentives?”, IZA Discussion Paper, No 2245.
Stevens, M. (1996), “Transferable training and poaching externalities”, In Booth, A.L. and D.J. Snower
(eds.) Acquiring skills, Cambridge University Press.
Stevens, M. (2001), “Should fi rms be required to pay for vocational training”, Economic Journal,
Vol.111:743, pp. 485-505.
Stock, J.H. and D.A. Wise (1990), “Pension, the option value of work and retirement”, Econometrica
Vol. 58:5, pp. 1151-1180.
Stoiber, S.A. (1990), “Family Leave Entitlements in Europe: Lessons for the United States”,
Compensation and Benefi ts Management, VI.
Tam, M. (1997), “Part-time employment: A bridge or a trap?”, Aldershot, Avebury.
The Economic Bureau of the President (2006), “Immigration and the Spanish economy” http://
www.la-moncloa.es/NR/rdonlyres/88165A8C-E40D-4D7E-9468-FB0C45516947/81221/
InmigrationandSpanishEconomy.pdf
Vamvakidis, A (2008) “Regional Wage differentiation and Wage Bargaining Systems in the EU”,
IMF Working Paper, No 08/43.
Viitanen, T.K. (2005), “Cost of Childcare and Female Employment in the UK”, Labour, Vol. 19,
No S1, pp. 149-170.
120ECB
Occasional Paper No 87
June 2008
Wößmann, L. (2007), “International Evidence on School Competition, Autonomy and
Accountability: A Review”, Peabody Journal of Education, Vol. 82, No 2-3, pp.473-497.
Weichselbaumer, D. and R. Winter-Ebmer (2005), “A Meta-Analysis of the International Gender
Wage Gap”, Journal of Economic Surveys, 9 (3), pp. 479-511.
Welch, F. (1979), “Effects of Cohort Size on Earnings: The Baby Boom Babies’ Financial Bust”,
Journal of Political Economy, Vol. 87, No 5.
121ECB
Occasional Paper No 87
June 2008
EUROPEAN
CENTRAL BANK
OCCAS IONAL
PAPER SERIES
EUROPEAN CENTRAL BANK
OCCASIONAL PAPER SERIES SINCE 2007
55 “Globalisation and euro area trade: Interactions and challenges” by U. Baumann and F. di Mauro,
February 2007.
56 “Assessing fi scal soundness: Theory and practice” by N. Giammarioli, C. Nickel, P. Rother,
J.-P. Vidal, March 2007.
57 “Understanding price developments and consumer price indices in south-eastern Europe”
by S. Herrmann and E. K. Polgar, March 2007.
58 “Long-Term Growth Prospects for the Russian Economy” by R. Beck, A. Kamps and
E. Mileva, March 2007.
59 “The ECB Survey of Professional Forecasters (SPF) a review after eight years’ experience”,
by C. Bowles, R. Friz, V. Genre, G. Kenny, A. Meyler and T. Rautanen, April 2007.
60 “Commodity price fl uctuations and their impact on monetary and fi scal policies in Western and
Central Africa” by U. Böwer, A. Geis and A. Winkler, April 2007.
61 “Determinants of growth in the central and eastern European EU Member States – A production
function approach” by O. Arratibel, F. Heinz, R. Martin, M. Przybyla, L. Rawdanowicz,
R. Serafi ni and T. Zumer, April 2007.
62 “Infl ation-linked bonds from a Central Bank perspective” by J. A. Garcia and A. van Rixtel,
June 2007.
63 “Corporate fi nance in the euro area – including background material”, Task Force of the
Monetary Policy Committee of the European System of Central Banks, June 2007.
64 “The use of portfolio credit risk models in central banks”, Task Force of the Market Operations
Committee of the European System of Central Banks, July 2007.
65 “The performance of credit rating systems in the assessment of collateral used in Eurosystem
monetary policy operations” by F. Coppens, F. González and G. Winkler, July 2007.
66 “Structural reforms in EMU and the role of monetary policy – a survey of the literature”
by N. Leiner-Killinger, V. López Pérez, R. Stiegert and G. Vitale, July 2007.
67 “Towards harmonised balance of payments and international investment position statistics – the
experience of the European compilers” by J.-M. Israël and C. Sánchez Muñoz, July 2007.
68 “The securities custody industry” by D. Chan, F. Fontan, S. Rosati and D. Russo, August 2007.
69 “Fiscal policy in Mediterranean countries – Developments, structures and implications for
monetary policy” by M. Sturm and F. Gurtner, August 2007.
122ECB
Occasional Paper No 87
June 2008
70 “The search for Columbus’ egg: Finding a new formula to determine quotas at the IMF”
by M. Skala, C. Thimann and R. Wölfi nger, August 2007.
71 “The economic impact of the Single Euro Payments Area” by H. Schmiedel, August 2007.
72 “The role of fi nancial markets and innovation in productivity and growth in Europe”
by P. Hartmann, F. Heider, E. Papaioannou and M. Lo Duca, September 2007.
73 “Reserve accumulation: objective or by-product?” by J. O. de Beaufort Wijnholds and
L. Søndergaard, September 2007.
74 “Analysis of revisions to general economic statistics” by H. C. Dieden and A. Kanutin,
October 2007.
75 “The role of other fi nancial intermediaries in monetary and credit developments in the euro area”
edited by P. Moutot and coordinated by D. Gerdesmeier, A. Lojschová and J. von Landesberger,
October 2007.
76 “Prudential and oversight requirements for securities settlement a comparison of cpss-iosco”
by D. Russo, G. Caviglia, C. Papathanassiou and S. Rosati, November 2007.
77 “Oil market structure, network effects and the choice of currency for oil invoicing” by E. Mileva
and N. Siegfried, November 2007.
78 “A framework for assessing global imbalances” by T. Bracke, M. Bussière, M. Fidora and
R. Straub, January 2008.
79 “The working of the eurosystem: monetary policy preparations and decision-making – selected
issues” by P. Moutot, A. Jung and F. P. Mongelli, January 2008.
80 “China’s and India’s roles in global trade and fi nance: twin titans for the new millennium?”
by M. Bussière and A. Mehl, January 2008.
81 “Measuring Financial Integration in New EU Member States” by M. Baltzer, L. Cappiello,
R.A. De Santis, and S. Manganelli, January 2008.
82 “The Sustainability of China’s Exchange Rate Policy and Capital Account Liberalisation”
by L. Cappiello and G. Ferrucci, February 2008.
83 “The predictability of monetary policy” by T. Blattner, M. Catenaro, M. Ehrmann, R. Strauch
and J. Turunen, March 2008.
84 “Short-term forecasting of GDP using large monthly datasets: a pseudo real-time forecast
evaluation exercise” by G. Rünstler, K. Barhoumi, R. Cristadoro, A. Den Reijer, A. Jakaitiene,
P. Jelonek, A. Rua, K. Ruth, S. Benk and C. Van Nieuwenhuyze, May 2008.
85 “Benchmarking the Lisbon Strategy” by D. Ioannou, M. Ferdinandusse, M. Lo Duca, and
W. Coussens, June 2008.
123ECB
Occasional Paper No 87
June 2008
EUROPEAN
CENTRAL BANK
OCCAS IONAL
PAPER SERIES
86 “Real convergence and the determinants of growth in EU candidate and potential candidate
countries: a panel data approach” by M. M. Borys, É. K. Polgár and A. Zlate, June 2008.
87 “Labour supply and employment in the euro area countries: developments and challenges”,
by a Task Force of the Monetary Policy Committee of the European System Of Central Banks,
June 2008.